release: v1.31.5

- Version bump 1.31.4 → 1.31.5 - server.json identifier URL updated (SHA256 set by CI after build) - CHANGELOG.md updated
feat: add check_update and update tools with background auto-check
2026-04-22 07:34:18 +07:00 · 2026-04-22 07:34:12 +07:00 · 2026-04-22 07:24:03 +07:00 · 2026-04-22 07:24:03 +07:00 · 2026-04-22 00:21:57 +00:00 · 2026-04-22 07:20:25 +07:00
39 changed files with 6424 additions and 842 deletions
@@ -0,0 +1,399 @@
+name: Build and Release
+
+on:
+  push:
+    tags:
+      - 'v*'
+  workflow_dispatch:
+    inputs:
+      version:
+        description: 'Version tag (e.g., v1.32.0)'
+        required: true
+        type: string
+
+jobs:
+  # ── Build binaries ───────────────────────────────────────────────────────────
+
+  build-macos:
+    runs-on: macos-latest
+    steps:
+      - uses: actions/checkout@v4
+
+      - name: Install Rust
+        uses: dtolnay/rust-toolchain@stable
+
+      - name: Cache Cargo
+        uses: actions/cache@v4
+        with:
+          path: |
+            ~/.cargo/registry
+            ~/.cargo/git
+            target
+          key: macos-cargo-${{ hashFiles('Cargo.lock') }}
+          restore-keys: macos-cargo-
+
+      - name: Build release binary
+        run: cargo build --release
+
+      - name: Package binary
+        run: |
+          mkdir -p dist/mcp-mt5-quant-macos
+          cp target/release/mt5-quant dist/mcp-mt5-quant-macos/
+          cp -r config dist/mcp-mt5-quant-macos/
+          cp README.md dist/mcp-mt5-quant-macos/
+          cp -r docs dist/mcp-mt5-quant-macos/
+          cd dist
+          tar -czf mcp-mt5-quant-macos-arm64.tar.gz mcp-mt5-quant-macos
+
+      - name: Upload artifact
+        uses: actions/upload-artifact@v4
+        with:
+          name: macos-binary
+          path: dist/mcp-mt5-quant-macos-arm64.tar.gz
+
+  build-linux:
+    runs-on: ubuntu-latest
+    steps:
+      - uses: actions/checkout@v4
+
+      - name: Install Rust
+        uses: dtolnay/rust-toolchain@stable
+
+      - name: Cache Cargo
+        uses: actions/cache@v4
+        with:
+          path: |
+            ~/.cargo/registry
+            ~/.cargo/git
+            target
+          key: linux-cargo-${{ hashFiles('Cargo.lock') }}
+          restore-keys: linux-cargo-
+
+      - name: Build release binary
+        run: cargo build --release
+
+      - name: Package binary
+        run: |
+          mkdir -p dist/mcp-mt5-quant-linux
+          cp target/release/mt5-quant dist/mcp-mt5-quant-linux/
+          cp -r config dist/mcp-mt5-quant-linux/
+          cp README.md dist/mcp-mt5-quant-linux/
+          cp -r docs dist/mcp-mt5-quant-linux/
+          cd dist
+          tar -czf mcp-mt5-quant-linux-x64.tar.gz mcp-mt5-quant-linux
+
+      - name: Upload artifact
+        uses: actions/upload-artifact@v4
+        with:
+          name: linux-binary
+          path: dist/mcp-mt5-quant-linux-x64.tar.gz
+
+  # ── GitHub Release ───────────────────────────────────────────────────────────
+
+  release:
+    needs: [build-macos, build-linux]
+    runs-on: ubuntu-latest
+    permissions:
+      contents: write
+    env:
+      RELEASE_TAG: ${{ github.event.inputs.version || github.ref_name }}
+    steps:
+      - uses: actions/checkout@v4
+
+      - name: Download macOS artifact
+        uses: actions/download-artifact@v4
+        with:
+          name: macos-binary
+          path: dist
+
+      - name: Download Linux artifact
+        uses: actions/download-artifact@v4
+        with:
+          name: linux-binary
+          path: dist
+
+      - name: Create GitHub Release
+        uses: softprops/action-gh-release@v2
+        with:
+          tag_name: ${{ env.RELEASE_TAG }}
+          name: ${{ env.RELEASE_TAG }}
+          files: |
+            dist/mcp-mt5-quant-macos-arm64.tar.gz
+            dist/mcp-mt5-quant-linux-x64.tar.gz
+          draft: false
+          prerelease: false
+          generate_release_notes: true
+        env:
+          GITHUB_TOKEN: ${{ secrets.GITHUB_TOKEN }}
+
+  # ── MCP Package ──────────────────────────────────────────────────────────────
+  # Builds the installable MCP package (binary + config + docs),
+  # computes its SHA256, and uploads to the release.
+
+  build-mcp-package:
+    needs: release
+    runs-on: macos-latest
+    permissions:
+      contents: write
+    outputs:
+      sha256: ${{ steps.compute-sha256.outputs.sha256 }}
+    env:
+      RELEASE_TAG: ${{ github.event.inputs.version || github.ref_name }}
+    steps:
+      - uses: actions/checkout@v4
+
+      - name: Download macOS binary
+        uses: actions/download-artifact@v4
+        with:
+          name: macos-binary
+          path: dist
+
+      - name: Build MCP package
+        run: |
+          mkdir -p mcp-package/mcp-server/bin
+          cd dist && tar -xzf mcp-mt5-quant-macos-arm64.tar.gz && cd ..
+          cp dist/mcp-mt5-quant-macos/mt5-quant mcp-package/mcp-server/bin/
+          cp -r config mcp-package/
+          cp -r docs mcp-package/
+          cp README.md mcp-package/
+          cp server.json mcp-package/
+          cd mcp-package && tar -czf ../mcp-mt5-quant-macos-arm64.tar.gz . && cd ..
+
+      - name: Compute SHA256
+        id: compute-sha256
+        run: |
+          SHA256=$(shasum -a 256 mcp-mt5-quant-macos-arm64.tar.gz | awk '{print $1}')
+          echo "sha256=$SHA256" >> "$GITHUB_OUTPUT"
+          echo "MCP Package SHA256: $SHA256"
+
+      - name: Upload MCP package to release
+        uses: softprops/action-gh-release@v2
+        with:
+          tag_name: ${{ env.RELEASE_TAG }}
+          files: mcp-mt5-quant-macos-arm64.tar.gz
+        env:
+          GITHUB_TOKEN: ${{ secrets.GITHUB_TOKEN }}
+
+      - name: Upload MCP package artifact
+        uses: actions/upload-artifact@v4
+        with:
+          name: mcp-package
+          path: mcp-mt5-quant-macos-arm64.tar.gz
+
+  # ── Update server.json SHA256 ────────────────────────────────────────────────
+  # Commits the computed SHA256 back to master so server.json always
+  # reflects the latest published package hash.
+
+  update-server-json:
+    needs: build-mcp-package
+    runs-on: ubuntu-latest
+    permissions:
+      contents: write
+    steps:
+      - uses: actions/checkout@v4
+        with:
+          ref: master
+          fetch-depth: 1
+          token: ${{ secrets.GITHUB_TOKEN }}
+
+      - name: Patch server.json
+        env:
+          PKG_SHA256: ${{ needs.build-mcp-package.outputs.sha256 }}
+          RELEASE_TAG: ${{ github.event.inputs.version || github.ref_name }}
+        run: |
+          VERSION="${RELEASE_TAG#v}"
+          echo "VERSION=$VERSION" >> "$GITHUB_ENV"
+          VERSION="$VERSION" PKG_SHA256="$PKG_SHA256" python3 - <<'PYEOF'
+          import json, os, re
+
+          sha256  = os.environ['PKG_SHA256']
+          version = os.environ['VERSION']
+
+          with open('server.json') as f:
+              data = json.load(f)
+
+          data['version'] = version
+          for pkg in data.get('packages', []):
+              pkg['version']    = version
+              pkg['fileSha256'] = sha256
+              if 'identifier' in pkg:
+                  pkg['identifier'] = re.sub(
+                      r'/v[0-9]+\.[0-9]+\.[0-9]+/',
+                      f'/v{version}/',
+                      pkg['identifier']
+                  )
+
+          with open('server.json', 'w') as f:
+              json.dump(data, f, indent=2)
+              f.write('\n')
+
+          print(f"Patched: version={version}, sha256={sha256[:16]}...")
+          PYEOF
+
+      - name: Commit server.json
+        run: |
+          git config user.name  "github-actions[bot]"
+          git config user.email "github-actions[bot]@users.noreply.github.com"
+          git add server.json
+          if git diff --cached --quiet; then
+            echo "server.json unchanged — no commit needed"
+          else
+            git commit -m "ci: update server.json SHA256 for v${VERSION} [skip ci]"
+            git push origin master
+            echo "Committed server.json update"
+          fi
+
+  # ── MCP Registry Publish ─────────────────────────────────────────────────────
+
+  mcp-publish:
+    needs: [build-mcp-package, update-server-json]
+    runs-on: ubuntu-latest
+    permissions:
+      contents: read
+      id-token: write   # Required for GitHub OIDC
+    steps:
+      - uses: actions/checkout@v4
+        with:
+          ref: master   # Pull master so server.json has the correct SHA256
+
+      - name: Download MCP package
+        uses: actions/download-artifact@v4
+        with:
+          name: mcp-package
+          path: .
+
+      - name: Install mcp-publisher
+        id: install-publisher
+        run: |
+          # Fetch the download URL for linux_amd64 directly from the release assets API
+          # (asset name format changed in v1.6.0 — look it up rather than constructing it)
+          URL=$(curl -sf \
+            -H "Accept: application/vnd.github+json" \
+            "https://api.github.com/repos/modelcontextprotocol/registry/releases/latest" \
+            | python3 -c "
+          import sys, json
+          r = json.load(sys.stdin)
+          tag = r['tag_name']
+          assets = r.get('assets', [])
+          match = next((a['browser_download_url'] for a in assets
+                        if 'linux' in a['name'] and 'amd64' in a['name']
+                        and a['name'].endswith('.tar.gz')
+                        and 'sbom' not in a['name']
+                        and 'sigstore' not in a['name']), None)
+          if match:
+              print(match)
+          else:
+              print(f'https://github.com/modelcontextprotocol/registry/releases/download/{tag}/mcp-publisher_linux_amd64.tar.gz', file=sys.stderr)
+              sys.exit(1)
+          " 2>/dev/null)
+
+          if [[ -z "$URL" ]]; then
+            echo "installed=false" >> "$GITHUB_OUTPUT"
+            echo "Could not resolve mcp-publisher download URL"
+          else
+            echo "Downloading: $URL"
+            if curl -fsSL -o mcp-publisher.tar.gz "$URL"; then
+              tar -xzf mcp-publisher.tar.gz 2>/dev/null || true
+              BINARY=$(find . -maxdepth 3 -name "mcp-publisher" ! -path "./.git/*" | head -1)
+              if [[ -n "$BINARY" && -f "$BINARY" ]]; then
+                chmod +x "$BINARY"
+                sudo mv "$BINARY" /usr/local/bin/mcp-publisher
+                echo "installed=true" >> "$GITHUB_OUTPUT"
+                echo "Installed: $(mcp-publisher --version 2>/dev/null || echo 'ok')"
+              else
+                echo "installed=false" >> "$GITHUB_OUTPUT"
+                echo "Binary not found in archive"
+              fi
+              rm -f mcp-publisher.tar.gz
+            else
+              echo "installed=false" >> "$GITHUB_OUTPUT"
+              echo "Download failed: $URL"
+            fi
+          fi
+
+      - name: Validate server.json before publish
+        id: validate
+        run: |
+          python3 - <<'PYEOF'
+          import json, sys
+
+          with open('server.json') as f:
+              data = json.load(f)
+
+          print(f"  name:    {data.get('name')}")
+          print(f"  version: {data.get('version')}")
+
+          ok = True
+          for pkg in data.get('packages', []):
+              sha = pkg.get('fileSha256', '')
+              ident = pkg.get('identifier', '')
+              print(f"  sha256:  {sha}")
+              print(f"  url:     {ident}")
+              if sha in ('', 'TBD_CI_WILL_UPDATE', 'TBD_UPDATED_BY_CI'):
+                  print('ERROR: fileSha256 is a placeholder — publish aborted')
+                  ok = False
+
+          sys.exit(0 if ok else 1)
+          PYEOF
+
+      - name: Publish to MCP Registry
+        if: steps.install-publisher.outputs.installed == 'true' && steps.validate.outcome == 'success'
+        run: |
+          echo "=== Authenticating with GitHub OIDC ==="
+          mcp-publisher login github-oidc || {
+            echo "OIDC login failed"
+            exit 1
+          }
+
+          echo "=== Publishing ==="
+          mcp-publisher publish && echo "✓ Published to MCP Registry" || {
+            echo ""
+            echo "══════════════════════════════════════════════════════════"
+            echo "  AUTOMATED MCP PUBLISH FAILED — manual steps:"
+            echo "══════════════════════════════════════════════════════════"
+            echo "  1. Download: https://github.com/modelcontextprotocol/registry/releases"
+            echo "  2. ./mcp-publisher login github"
+            echo "  3. ./mcp-publisher publish"
+            echo "  Web UI: https://registry.modelcontextprotocol.io"
+            echo "══════════════════════════════════════════════════════════"
+            exit 0   # Don't fail the workflow — release succeeded
+          }
+
+      - name: Skip notice
+        if: steps.install-publisher.outputs.installed != 'true'
+        run: |
+          echo "mcp-publisher not installed — skipping automated publish"
+          echo "Manual: https://registry.modelcontextprotocol.io"
+
+  # ── Release Summary ──────────────────────────────────────────────────────────
+
+  release-info:
+    needs: [mcp-publish]
+    runs-on: ubuntu-latest
+    env:
+      RELEASE_TAG: ${{ github.event.inputs.version || github.ref_name }}
+    steps:
+      - name: Summary
+        run: |
+          echo ""
+          echo "╔══════════════════════════════════════════════════════════════╗"
+          echo "║              Release ${RELEASE_TAG} complete!                "
+          echo "╚══════════════════════════════════════════════════════════════╝"
+          echo ""
+          echo "GitHub Release:"
+          echo "  https://github.com/masdevid/mt5-quant/releases/tag/${RELEASE_TAG}"
+          echo ""
+          echo "MCP Registry: io.github.masdevid/mt5-quant"
+          echo ""
+          echo "── MCP Client Registration ──────────────────────────────────"
+          echo ""
+          echo "Claude Code:"
+          echo "  claude mcp add io.github.masdevid/mt5-quant -- ~/.local/bin/mt5-quant"
+          echo ""
+          echo "Windsurf  (~/.codeium/windsurf/mcp_config.json):"
+          echo '  {"mcpServers": {"io.github.masdevid/mt5-quant": {"command": "~/.local/bin/mt5-quant"}}}'
+          echo ""
+          echo "Cursor / VS Code  (~/.cursor/mcp.json):"
+          echo '  {"mcpServers": {"mt5-quant": {"command": "~/.local/bin/mt5-quant"}}}'
+          echo ""
+          echo "─────────────────────────────────────────────────────────────"
@@ -0,0 +1,22 @@
+name: Rust
+
+on:
+  push:
+    branches: [ "master" ]
+  pull_request:
+    branches: [ "master" ]
+
+env:
+  CARGO_TERM_COLOR: always
+
+jobs:
+  build:
+
+    runs-on: ubuntu-latest
+
+    steps:
+    - uses: actions/checkout@v4
+    - name: Build
+      run: cargo build --verbose
+    - name: Run tests
+      run: cargo test --verbose
@@ -20,14 +20,12 @@ config/baseline.json
 config/backtest_history.json

 # Claude Code (personal, not for repo)
-CLAUDE.md
+/CLAUDE.md
+.claude/

 # Windsurf (local workflows, not for repo)
 .windsurf/

-# GitHub Actions (local workflows, not for repo)
-.github/workflows/
-
 # macOS
 .DS_Store

@@ -0,0 +1,42 @@
+# Changelog
+
+## [1.31.5] — 2026-04-22
+
+- feat: add check_update and update tools with background auto-check
+- release: v1.31.4
+- fix: update all scripts for correctness and consistency
+- release: v1.31.3
+- docs: clean up public repo — remove IDE files, fix stale refs
+- release: v1.31.2
+- refactor: reduce handler boilerplate in analysis and experts modules
+- fix: registryType mcpbPackageType → mcpb
+
+
+## [1.31.4] — 2026-04-22
+
+- fix: update all scripts for correctness and consistency
+- release: v1.31.3
+- docs: clean up public repo — remove IDE files, fix stale refs
+- release: v1.31.2
+- refactor: reduce handler boilerplate in analysis and experts modules
+- fix: registryType mcpbPackageType → mcpb
+
+
+## [1.31.3] — 2026-04-22
+
+- docs: clean up public repo — remove IDE files, fix stale refs
+- release: v1.31.2
+- refactor: reduce handler boilerplate in analysis and experts modules
+- fix: registryType mcpbPackageType → mcpb
+
+
+## [1.31.2] — 2026-04-22
+
+- refactor: reduce handler boilerplate in analysis and experts modules
+- fix: registryType mcpbPackageType → mcpb
+
+
+## [1.31.1] — 2026-04-22
+
+- Minor improvements and bug fixes
+
@@ -481,7 +481,7 @@ dependencies = [

 [[package]]
 name = "mt5-quant"
-version = "1.28.0"
+version = "1.31.5"
 dependencies = [
 "anyhow",
 "base64",
@@ -1,6 +1,6 @@
 [package]
 name = "mt5-quant"
-version = "1.28.0"
+version = "1.31.5"
 edition = "2021"
 description = "MT5-Quant MCP Server - Exposes MT5 backtest and optimization tools via MCP"
 authors = ["masdevid <masdevid@example.com>"]
@@ -1,6 +1,6 @@
 # MT5-Quant

-**MCP server for MT5 strategy development on macOS/Linux.** 43 tools to compile, backtest, analyze, and optimize MQL5 Expert Advisors — no Windows required.
+**MCP server for MT5 strategy development on macOS/Linux.** 87 tools to compile, backtest, analyze, optimize, debug crashes, and manage MQL5 Expert Advisors — no Windows required.

 ```
 You: "Backtest MyEA Jan-Mar, what caused the February drawdown?"
@@ -20,17 +20,29 @@ Claude: [compile → clean → backtest → analyze 1,847 deals]
 | MQL5 compilation | ✅ | ❌ | ❌ |
 | macOS/Linux native | ✅ | Windows only | Cloud |
 | Optimization | ✅ Background | ❌ | ✅ Paid |
+| Crash debugging | ✅ Wine/MT5 diagnostics | ❌ | ❌ |

 ## Quick Install

+### 1. Download & Setup
+
 ```bash
-curl -L -o mt5.tar.gz https://github.com/masdevid/mt5-mcp/releases/latest/download/mt5-quant-macos-arm64.tar.gz
+curl -L -o mt5.tar.gz https://github.com/masdevid/mt5-quant/releases/latest/download/mt5-quant-macos-arm64.tar.gz
 tar -xzf mt5.tar.gz
 bash scripts/setup.sh
-claude mcp add MT5-Quant -- $(pwd)/mt5-quant
 ```

-**[Full Setup →](docs/QUICKSTART.md)**
+### 2. Register MCP Server
+
+| Platform | Command / Config | Docs |
+|----------|------------------|------|
+| **Claude Code** | `claude mcp add mt5-quant -- $(pwd)/mt5-quant` | [Setup →](docs/QUICKSTART.md) |
+| **Windsurf** | Edit `~/.codeium/windsurf/mcp_config.json` | [WINDSURF.md →](docs/WINDSURF.md) |
+| **Cursor** | Edit `~/.cursor/mcp.json` or use Settings → MCP | [CURSOR.md →](docs/CURSOR.md) |
+| **VS Code** | Edit `.vscode/mcp.json` or run `MCP: Add Server` | [VSCODE.md →](docs/VSCODE.md) |
+| **Claude Desktop** | Agent Panel → ... → MCP Servers → Edit configuration | [CLAUDE.md →](docs/CLAUDE.md) |
+
+> **Note:** Use absolute paths like `/Users/name/mt5-quant/mt5-quant` or `$(pwd)/mt5-quant`, not relative paths like `./mt5-quant`.

 ## Quick Start

@@ -45,20 +57,27 @@ The AI runs the full pipeline: compile → clean cache → backtest → extract
 | Doc | Purpose |
 |-----|---------|
 | [QUICKSTART.md](docs/QUICKSTART.md) | Complete setup for macOS/Linux |
+| [WINDSURF.md](docs/WINDSURF.md) | Windsurf IDE setup |
+| [CURSOR.md](docs/CURSOR.md) | Cursor IDE setup |
+| [VSCODE.md](docs/VSCODE.md) | VS Code setup |
+| [CLAUDE.md](docs/CLAUDE.md) | Claude Desktop setup |
 | [CONFIG.md](docs/CONFIG.md) | Configuration reference |
-| [WINDSURF.md](docs/WINDSURF.md) | Windsurf IDE integration |
-| [TOOLS.md](docs/MCP_TOOLS.md) | All 43 tools (31 documented) |
+| [TOOLS.md](docs/MCP_TOOLS.md) | All 87 tools documented |
 | [ARCHITECTURE.md](docs/ARCHITECTURE.md) | Design and internals |
 | [TROUBLESHOOTING.md](docs/TROUBLESHOOTING.md) | Common issues |
 | [REMOTE_AGENTS.md](docs/REMOTE_AGENTS.md) | Linux optimization agents |

-## MCP Tools (43)
+## MCP Tools (89)

 ### Core workflow

 | Tool | Description |
 |------|-------------|
 | `run_backtest` | Full pipeline: compile → clean → backtest → extract → analyze |
+| `run_backtest_quick` | Quick backtest using pre-compiled EA (skip compile) |
+| `run_backtest_only` | Backtest only - just extract raw trades, no analysis |
+| `launch_backtest` | Fire-and-forget: launch MT5 backtest, poll for completion |
+| `get_backtest_status` | Poll running backtest status (MT5 running, report found, elapsed time) |
 | `run_optimization` | Genetic optimization (background, returns immediately) |
 | `get_optimization_results` | Parse optimization results after MT5 finishes |
 | `analyze_report` | Read `analysis.json` from any report directory |
@@ -84,12 +103,26 @@ The AI runs the full pipeline: compile → clean cache → backtest → extract

 Use these for targeted analysis, or `analyze_report` to run all at once.

+### Deal-Level Analytics (New)
+
+| Tool | Description |
+|------|-------------|
+| `list_deals` | List individual deals with filters (type, profit range, volume, dates) |
+| `search_deals_by_comment` | Full-text search in deal comments (e.g., "Layer #3") |
+| `search_deals_by_magic` | Filter deals by EA magic number |
+| `analyze_profit_distribution` | Profit histogram: small/medium/large wins and losses |
+| `analyze_time_performance` | Performance by hour of day and day of week |
+| `analyze_hold_time_distribution` | Hold time buckets + correlation with profit |
+| `analyze_layer_performance` | Grid/martingale layer analysis from comments |
+| `analyze_volume_vs_profit` | Volume correlation + performance by lot size |
+| `analyze_costs` | Commission and swap impact on profitability |
+| `analyze_efficiency` | Profit per hour/day, annualized return, trade frequency |
+
 ### Monitoring

 | Tool | Description |
 |------|-------------|
 | `verify_setup` | Check Wine/MT5 paths, Wine version, and EA/set file counts |
-| `get_backtest_status` | Check live progress of a running backtest pipeline |
 | `get_optimization_status` | Check live state of a background optimization job |
 | `list_jobs` | All optimization jobs with compact status in one call |

@@ -100,6 +133,14 @@ Use these for targeted analysis, or `analyze_report` to run all at once.
 | `list_reports` | Compact table of all runs with key metrics — no full analysis needed |
 | `get_latest_report` | Get most recent report with optional equity chart |
 | `search_reports` | Find reports by EA, symbol, date range, or profit criteria |
+| `get_report_by_id` | Get specific report by ID with equity chart |
+| `get_reports_summary` | Aggregate stats: counts, averages, pass rates |
+| `get_best_reports` | Top N reports sorted by any metric (profit factor, drawdown, etc.) |
+| `search_reports_by_tags` | Find reports by tags |
+| `search_reports_by_date_range` | Query by backtest date range |
+| `search_reports_by_notes` | Full-text search in report notes |
+| `get_reports_by_set_file` | Find all reports using a specific .set file |
+| `get_comparable_reports` | Find comparable reports (same EA/symbol/timeframe) |
 | `tail_log` | Read last N lines of any log; `filter=errors` to see only failures |
 | `prune_reports` | Delete old report directories, keep last N (skips `_opt` dirs) |

@@ -120,6 +161,44 @@ Use these for targeted analysis, or `analyze_report` to run all at once.
 | `cache_status` | MT5 tester cache size breakdown by symbol — check before cleaning |
 | `clean_cache` | Delete tester cache files; supports per-symbol and `dry_run` |

+### Pre-flight & Validation
+
+| Tool | Description |
+|------|-------------|
+| `get_active_account` | Get current MT5 account session (login, server, available symbols) |
+| `check_symbol_data_status` | Validate symbol has sufficient history data for date range |
+| `check_mt5_status` | Check if MT5 terminal is installed and ready |
+| `validate_ea_syntax` | Pre-compile syntax check without running full compilation |
+
+### Debugging & Diagnostics (New)
+
+| Tool | Description |
+|------|-------------|
+| `diagnose_wine` | Check Wine installation, version, and prefix health |
+| `get_mt5_logs` | Get MT5 terminal, tester, or MetaEditor logs with filtering |
+| `search_mt5_errors` | Search logs for error patterns (crash, exception, access violation) |
+| `check_mt5_process` | Check if MT5 processes are running, get PID, CPU, memory usage |
+| `kill_mt5_process` | Kill stuck MT5 processes (force=true for wineserver) |
+| `check_system_resources` | Check disk space, memory, CPU availability |
+| `validate_mt5_config` | Validate terminal.ini and tester configuration files |
+| `get_wine_prefix_info` | Get Wine prefix details: Windows version, installed programs, registry |
+| `get_backtest_crash_info` | Investigate backtest failures: incomplete markers, missing deals.csv, errors |
+
+### Project Management
+
+| Tool | Description |
+|------|-------------|
+| `init_project` | Scaffold new MQL5 project with templates (scalper/swing/grid/basic) |
+| `create_set_template` | Generate .set parameter file from EA input variables |
+| `export_report` | Export backtest report to CSV, JSON, or Markdown |
+
+### History & Comparison
+
+| Tool | Description |
+|------|-------------|
+| `get_backtest_history` | List all backtests for EA/symbol with summary metrics |
+| `compare_backtests` | Compare 2+ backtest results side-by-side with analysis |
+
 ### .set file — read / write

 | Tool | Description |
@@ -138,12 +217,28 @@ Use these for targeted analysis, or `analyze_report` to run all at once.
 | `diff_set_files` | Side-by-side diff of two `.set` files — only changed params returned |
 | `set_from_optimization` | Generate a clean backtest `.set` from `get_optimization_results` params; optionally narrow sweep |

+### Search & Discovery
+
+| Tool | Description |
+|------|-------------|
+| `search_experts` | Search EAs by name pattern across all directories |
+| `search_indicators` | Search indicators by name pattern |
+| `search_scripts` | Search scripts by name pattern |
+| `copy_indicator_to_project` | Copy indicator to project directory |
+| `copy_script_to_project` | Copy script to project directory |
+
 Full schema: [docs/MCP_TOOLS.md](docs/MCP_TOOLS.md)

 ## Troubleshooting

 Run `verify_setup` from Claude first — it checks all paths and returns actionable hints.

+For crashes or unexplained failures during backtest/compile/optimization:
+- `diagnose_wine` — Check Wine installation and prefix health
+- `search_mt5_errors` — Find crash causes in logs
+- `check_mt5_process` + `kill_mt5_process` — Detect and kill stuck processes
+- `get_backtest_crash_info` — Investigate failed backtest reports
+
 **[Full Troubleshooting Guide →](docs/TROUBLESHOOTING.md)**

 ---
@@ -153,5 +248,3 @@ Run `verify_setup` from Claude first — it checks all paths and returns actiona
 MIT

 ---
-
-*Built from battle-tested production infrastructure. Every edge case in the pipeline was hit in production.*
@@ -0,0 +1,138 @@
+# Claude Desktop MCP Integration Setup
+
+## Quick Setup
+
+### Option 1: Install via MCP Registry (Recommended)
+
+Search for `mt5-quant` in Claude Desktop's MCP manager, or add directly to your config.
+
+### Option 2: Download Prebuilt Binary
+
+```bash
+# macOS (Apple Silicon)
+curl -L -o mt5-quant.tar.gz https://github.com/masdevid/mt5-quant/releases/latest/download/mt5-quant-macos-arm64.tar.gz
+tar -xzf mt5-quant.tar.gz
+
+# Linux (x64)
+curl -L -o mt5-quant.tar.gz https://github.com/masdevid/mt5-quant/releases/latest/download/mt5-quant-linux-x64.tar.gz
+tar -xzf mt5-quant.tar.gz
+```
+
+### Option 3: Build from Source
+
+```bash
+git clone https://github.com/masdevid/mt5-quant
+cd mt5-quant
+cargo build --release
+```
+
+## Configure Claude Desktop
+
+### Step 1: Open MCP Configuration
+
+Edit `~/Library/Application Support/Claude/claude_desktop_config.json` (macOS) or `%APPDATA%\Claude\claude_desktop_config.json` (Windows).
+
+### Step 2: Add mt5-quant
+
+```json
+{
+  "mcpServers": {
+    "mt5-quant": {
+      "command": "/absolute/path/to/mt5-quant"
+    }
+  }
+}
+```
+
+### Step 3: Reload and Verify
+
+1. Save the file
+2. **Restart Claude Desktop**
+3. In a new conversation, type: `What tools do you have access to?`
+4. The agent should list MT5 tools like `verify_setup`, `run_backtest`, etc.
+
+## Full Example Configuration
+
+```json
+{
+  "mcpServers": {
+    "mt5-quant": {
+      "command": "/Users/name/mt5-quant/target/release/mt5-quant"
+    },
+    "github": {
+      "command": "npx",
+      "args": ["-y", "@modelcontextprotocol/server-github"],
+      "env": {
+        "GITHUB_PERSONAL_ACCESS_TOKEN": "${env:GITHUB_TOKEN}"
+      }
+    }
+  }
+}
+```
+
+## Environment Variables
+
+Claude Desktop supports `${env:VAR_NAME}` syntax for environment variable substitution:
+
+```json
+{
+  "mcpServers": {
+    "mt5-quant": {
+      "command": "/Users/name/mt5-quant/target/release/mt5-quant",
+      "env": {
+        "MT5_MCP_HOME": "${env:HOME}/.config/mt5-quant"
+      }
+    }
+  }
+}
+```
+
+## Verify Setup
+
+In a Claude Desktop conversation, type:
+
+```
+Run verify_setup
+```
+
+Expected output:
+```
+Wine:    /Applications/MetaTrader 5.app/.../wine64
+MT5 dir: ~/Library/Application Support/.../MetaTrader 5
+Display: gui
+Arch:    arch -x86_64
+```
+
+## Troubleshooting
+
+### "Tool not found" or server not appearing
+
+1. Double-check the absolute path in `claude_desktop_config.json`
+2. Ensure the binary has execute permissions: `chmod +x /path/to/mt5-quant`
+3. Restart Claude Desktop completely
+4. Check Claude Desktop logs: `~/Library/Logs/Claude/` (macOS)
+
+### JSON syntax errors
+
+1. Validate the JSON at [jsonlint.com](https://jsonlint.com)
+2. Ensure no trailing commas
+3. Use absolute paths (not `~` or relative paths)
+
+### Agent hallucinating tool parameters
+
+1. Verify the tool is available: "List your available MCP tools"
+2. Start a new conversation
+3. Be explicit: "Use the `run_backtest` tool with expert=MyEA"
+
+## Configuration Location
+
+| Platform | Path |
+|----------|------|
+| macOS | `~/Library/Application Support/Claude/claude_desktop_config.json` |
+| Windows | `%APPDATA%\Claude\claude_desktop_config.json` |
+
+## Resources
+
+- [Claude Desktop MCP Documentation](https://docs.anthropic.com/en/docs/claude-code/mcp)
+- [MCP Server Reference](https://github.com/modelcontextprotocol/servers)
+- [MT5-Quant Tools Reference](./MCP_TOOLS.md)
@@ -0,0 +1,129 @@
+# Cursor MCP Integration Setup
+
+## Quick Setup
+
+### Option 1: Download Prebuilt Binary (Recommended)
+
+```bash
+# macOS (Apple Silicon)
+curl -L -o mt5-quant.tar.gz https://github.com/masdevid/mt5-quant/releases/latest/download/mt5-quant-macos-arm64.tar.gz
+tar -xzf mt5-quant.tar.gz
+
+# Linux (x64)
+curl -L -o mt5-quant.tar.gz https://github.com/masdevid/mt5-quant/releases/latest/download/mt5-quant-linux-x64.tar.gz
+tar -xzf mt5-quant.tar.gz
+```
+
+### Option 2: Build from Source
+
+```bash
+cargo build --release
+```
+
+## Configure Cursor
+
+### Method 1: Settings UI (Recommended)
+
+1. Open Cursor Settings (`Cmd/Ctrl + ,`)
+2. Navigate to **Features** → **MCP**
+3. Click **Add Custom MCP**
+4. Enter:
+   - **Name**: `mt5-quant`
+   - **Command**: `~/.local/bin/mt5-quant`
+   - **Type**: `stdio`
+
+### Method 2: Edit mcp.json Directly
+
+Add to `~/.cursor/mcp.json` (global) or `.cursor/mcp.json` (project):
+
+```json
+{
+  "mcpServers": {
+    "mt5-quant": {
+      "type": "stdio",
+      "command": "~/.local/bin/mt5-quant"
+    }
+  }
+}
+```
+
+Create the file if it doesn't exist:
+
+```bash
+mkdir -p ~/.cursor
+cat > ~/.cursor/mcp.json << 'EOF'
+{
+  "mcpServers": {
+    "mt5-quant": {
+      "type": "stdio",
+      "command": "/path/to/mt5-quant"
+    }
+  }
+}
+EOF
+```
+
+## Verify Setup
+
+In Cursor chat, type:
+
+```
+Run verify_setup
+```
+
+Expected output:
+```
+Wine:    /Applications/MetaTrader 5.app/.../wine64
+MT5 dir: ~/Library/Application Support/.../MetaTrader 5
+Display: gui
+Arch:    arch -x86_64
+```
+
+## Configuration Locations
+
+| Scope | Path | Use Case |
+|-------|------|----------|
+| Global | `~/.cursor/mcp.json` | Available in all projects |
+| Project | `.cursor/mcp.json` | Project-specific tools |
+
+## Troubleshooting
+
+### MCP server not appearing
+
+1. Check MCP panel in Cursor Settings
+2. Verify the path is absolute (not relative)
+3. Test binary: `/path/to/mt5-quant --help`
+4. View MCP logs: Output panel → select "MCP" from dropdown
+
+### Config interpolation
+
+Cursor supports variable substitution in `mcp.json`:
+
+```json
+{
+  "mcpServers": {
+    "mt5-quant": {
+      "command": "${userHome}/bin/mt5-quant",
+      "args": ["--config", "${workspaceFolder}/config.yaml"]
+    }
+  }
+}
+```
+
+Available variables:
+- `${userHome}` - Home directory
+- `${workspaceFolder}` - Project root
+- `${workspaceFolderBasename}` - Project folder name
+- `${env:VAR_NAME}` - Environment variable
+
+### Tool not found errors
+
+If the agent says "Tool not found":
+1. Check the server is enabled in MCP settings
+2. Try disabling and re-enabling the server
+3. Restart Cursor
+
+## Resources
+
+- [Cursor MCP Documentation](https://cursor.com/docs/context/mcp)
+- [MT5-Quant Tools Reference](./MCP_TOOLS.md)
@@ -2,11 +2,15 @@

 Full input/output schemas for MT5-Quant tools.

-> **Documentation Status:** This file documents 31 of 43 total tools. Missing:
+> **Documentation Status:** This file documents 56 of 89 total tools. Missing:
 > - `list_experts`, `list_indicators`, `list_scripts`
-> - `healthcheck`
-> - `search_reports`, `get_latest_report`
+> - `healthcheck`, `list_symbols`
+> - Reports query: `search_reports`, `get_latest_report`, `list_reports`, `prune_reports`, `tail_log`, `get_report_by_id`, `get_reports_summary`, `get_best_reports`, `search_reports_by_tags`, `search_reports_by_date_range`, `search_reports_by_notes`, `get_reports_by_set_file`, `get_comparable_reports`
 > - Granular analytics: `analyze_monthly_pnl`, `analyze_drawdown_events`, `analyze_top_losses`, `analyze_loss_sequences`, `analyze_position_pairs`, `analyze_direction_bias`, `analyze_streaks`, `analyze_concurrent_peak`
+> - Deal analytics: `list_deals`, `search_deals_by_comment`, `search_deals_by_magic`, `analyze_profit_distribution`, `analyze_time_performance`, `analyze_hold_time_distribution`, `analyze_layer_performance`, `analyze_volume_vs_profit`, `analyze_costs`, `analyze_efficiency`
+> - Archive/history tools: `archive_report`, `archive_all_reports`, `get_history`, `annotate_history`, `promote_to_baseline`
+> - Experts search: `search_experts`, `search_indicators`, `search_scripts`, `copy_indicator_to_project`, `copy_script_to_project`
+> - Debugging/diagnostics: `diagnose_wine`, `get_mt5_logs`, `search_mt5_errors`, `check_mt5_process`, `kill_mt5_process`, `check_system_resources`, `validate_mt5_config`, `get_wine_prefix_info`, `get_backtest_crash_info`

 ---

@@ -141,6 +145,124 @@ Run a complete backtest pipeline: compile → clean cache → backtest → extra

 ---

+## `run_backtest_quick`
+
+Quick backtest using pre-compiled EA: clean cache → backtest → extract → analyze.
+
+**When to call:** When EA code hasn't changed and you just want to test different parameters or date ranges. Faster than `run_backtest` because it skips compilation.
+
+### Input schema
+
+Same as `run_backtest`, but `skip_compile` is automatically set to `true`.
+
+### Output schema
+
+Same as `run_backtest`.
+
+---
+
+## `run_backtest_only`
+
+Backtest only: clean cache → backtest → extract. No analysis phase.
+
+**When to call:** When you just need raw trade data (deals.csv) and don't need analytics. Fastest option for batch processing.
+
+### Input schema
+
+Same as `run_backtest`, but `skip_compile` and `skip_analyze` are automatically set to `true`.
+
+### Output schema
+
+Same as `run_backtest` but without `analysis_summary`.
+
+---
+
+## `launch_backtest`
+
+Fire-and-forget mode: compile → clean → launch MT5 backtest, return immediately with job info.
+
+**When to call:** When you want to launch a backtest without waiting for completion. Use `get_backtest_status` to poll for completion.
+
+### Input schema
+
+```typescript
+{
+  expert: string;          // Required. EA name without path or extension
+  symbol?: string;         // Trading symbol (default: from config or first available)
+  from_date?: string;      // Start date YYYY.MM.DD (default: past complete month)
+  to_date?: string;        // End date YYYY.MM.DD (default: past complete month)
+  timeframe?: string;      // M1, M5, M15, M30, H1, H4, D1 (default: M5)
+  deposit?: number;        // Initial deposit (default: 10000)
+  model?: 0 | 1 | 2;      // Tick model (default: 0)
+  set_file?: string;       // Path to .set parameter file
+  skip_compile?: boolean;  // Skip compilation
+  skip_clean?: boolean;    // Skip cache cleaning
+  timeout?: number;        // Max time in seconds (default: 900)
+  gui?: boolean;          // Enable MT5 visualization
+}
+```
+
+### Output schema
+
+```typescript
+{
+  success: true;
+  message: string;         // "Backtest launched successfully..."
+  report_id: string;       // e.g., "20250122_034455_MyEA_XAUUSD_M5"
+  report_dir: string;      // Full path to report directory
+  expert: string;
+  symbol: string;
+  timeframe: string;
+  launched_at: string;     // ISO8601 timestamp
+  timeout_seconds: number;
+  poll_hint: string;       // "Call get_backtest_status with report_dir to check progress"
+}
+```
+
+---
+
+## `get_backtest_status`
+
+Check progress of a running backtest pipeline launched via `launch_backtest`.
+
+**When to call:** Poll periodically after calling `launch_backtest` to check completion status.
+
+### Input schema
+
+```typescript
+{
+  report_dir: string;  // Report directory path from launch_backtest output
+}
+```
+
+### Output schema
+
+```typescript
+{
+  success: true;
+  report_dir: string;
+  status: "completed" | "running" | "failed" | "in_progress" | "not_started";
+  stage: string;           // Current pipeline stage: COMPILE, CLEAN, BACKTEST, EXTRACT, ANALYZE, DONE
+  is_complete: boolean;    // True if backtest finished successfully
+  mt5_running: boolean;    // Whether MT5 process is active
+  report_found: boolean;     // Whether report file exists
+  metrics_extracted: boolean;
+  deals_extracted: boolean;
+  elapsed_seconds: number;   // Time since launch
+  message: string;          // Human-readable status message
+  job?: {
+    report_id: string;
+    expert: string;
+    symbol: string;
+    timeframe: string;
+    launched_at: string;
+    timeout_seconds: number;
+  }
+}
+```
+
+---
+
 ## `run_optimization`

 Launch genetic parameter optimization as a detached background process.
@@ -1237,6 +1359,323 @@ List all `.set` files in the MT5 tester profiles directory with param counts, sw

 ---

+## `get_active_account`
+
+Get current MT5 account session information: login, server, and available symbols. This is essential for pre-flight checks to ensure symbol availability before backtesting.
+
+### Input schema
+
+```typescript
+{}  // No parameters
+```
+
+### Output schema
+
+```typescript
+{
+  success: boolean;
+  ready_for_backtest: boolean;    // true if account exists and symbols available
+  account: {
+    login: string;
+    server: string;
+  } | null;
+  server: string;                   // Active server name
+  available_servers: string[];      // All servers with history data
+  symbols: string[];                // Symbols available for active server
+  symbol_count: number;
+  hint: string;                     // "Ready for backtesting" or instructions
+}
+```
+
+---
+
+## `check_symbol_data_status`
+
+Validate if a symbol has sufficient historical tick data for a specified date range before running backtest. Prevents failed backtests due to missing history data.
+
+### Input schema
+
+```typescript
+{
+  symbol: string;        // e.g., "XAUUSDc"
+  from_date: string;     // "YYYY.MM.DD"
+  to_date: string;       // "YYYY.MM.DD"
+}
+```
+
+### Output schema
+
+```typescript
+{
+  success: boolean;
+  symbol: string;
+  server: string;
+  has_sufficient_data: boolean;
+  requested_range: { from: string; to: string };
+  data_range: string;           // "YYYY.MM.DD - YYYY.MM.DD" or "unknown"
+  years_available: number;      // Count of years with data
+  hcc_files_count: number;      // Number of history cache files
+  warnings: string[] | null;    // Data range issues
+  suggestion: string;           // Action recommendation
+}
+```
+
+---
+
+## `check_mt5_status`
+
+Check if MT5 terminal is properly installed and configured. Returns comprehensive status of all required components.
+
+### Input schema
+
+```typescript
+{}  // No parameters
+```
+
+### Output schema
+
+```typescript
+{
+  success: boolean;
+  terminal_ready: boolean;      // true if all components present
+  checks: {
+    mt5_dir_exists: boolean;
+    terminal64_exe: boolean;
+    metaeditor64_exe: boolean;
+    metatester64_exe: boolean;
+    wine_executable: boolean;
+    wine_path: string | null;
+  };
+  mt5_version: string | null;
+  current_account: {
+    login: string;
+    server: string;
+  } | null;
+  hint: string;
+}
+```
+
+---
+
+## `get_backtest_history`
+
+List all backtests previously run for a specific EA and/or symbol with summary metrics. Use for tracking performance over time.
+
+### Input schema
+
+```typescript
+{
+  expert?: string;       // Filter by EA name
+  symbol?: string;       // Filter by symbol
+  limit?: number;        // Max results (default: 10)
+}
+```
+
+### Output schema
+
+```typescript
+{
+  success: boolean;
+  count: number;
+  total: number;
+  filters: {
+    expert: string | null;
+    symbol: string | null;
+  };
+  history: Array<{
+    report_dir: string;
+    date: string | null;
+    expert: string | null;
+    symbol: string | null;
+    period: string | null;
+    profit: number | null;
+    profit_factor: number | null;
+    expected_payoff: number | null;
+    drawdown_pct: number | null;
+    total_trades: number | null;
+    win_rate: number | null;
+  }>;
+  hint: string;
+}
+```
+
+---
+
+## `compare_backtests`
+
+Compare two or more backtest results side-by-side with key metrics analysis. Includes profit/drawdown differences and verdict on which performed better.
+
+### Input schema
+
+```typescript
+{
+  report_dirs: string[];  // List of report directory paths to compare
+}
+```
+
+### Output schema
+
+```typescript
+{
+  success: boolean;
+  count: number;
+  comparisons: Array<{
+    report_dir: string;
+    expert: string | null;
+    symbol: string | null;
+    net_profit: number | null;
+    profit_factor: number | null;
+    drawdown_pct: number | null;
+    total_trades: number | null;
+    win_rate: number | null;
+    expected_payoff: number | null;
+    recovery_factor: number | null;
+    sharpe_ratio: number | null;
+  }>;
+  analysis: Array<{
+    compare_to: string | null;
+    report: string | null;
+    profit_diff: number;
+    profit_pct_change: number;
+    drawdown_diff: number;
+    profit_factor_diff: number;
+    verdict: "better" | "worse" | "mixed";
+  }> | null;
+  verdict: string | null;  // "Best: <report_dir>"
+}
+```
+
+---
+
+## `init_project`
+
+Create a new MQL5 project with standard directory structure and template files. Supports scalper, swing, grid, and basic templates.
+
+### Input schema
+
+```typescript
+{
+  name: string;                    // Project name (used for EA filename)
+  template?: "scalper" | "swing" | "grid" | "basic";  // Default: "basic"
+}
+```
+
+### Output schema
+
+```typescript
+{
+  success: boolean;
+  project_name: string;
+  template: string;
+  created_files: string[];  // Paths to created files
+  hint: string;
+}
+```
+
+---
+
+## `validate_ea_syntax`
+
+Perform pre-compile syntax check on MQL5 source file without running full compilation. Detects common issues before expensive MetaEditor compilation.
+
+### Input schema
+
+```typescript
+{
+  path: string;  // Path to .mq5 source file
+}
+```
+
+### Output schema
+
+```typescript
+{
+  success: boolean;
+  valid: boolean;
+  path: string;
+  checks: {
+    has_on_init: boolean;
+    has_on_tick: boolean;
+    has_on_deinit: boolean;
+    lines: number;
+  };
+  errors: Array<{
+    line: number;
+    message: string;
+    severity: "error";
+  }> | null;
+  warnings: Array<{
+    line: number;
+    message: string;
+    severity: "warning";
+  }> | null;
+  hint: string;
+}
+```
+
+---
+
+## `create_set_template`
+
+Generate a .set parameter file template based on an EA's input variables. Automatically parses input declarations from source code.
+
+### Input schema
+
+```typescript
+{
+  ea: string;              // EA name or path to .mq5/.ex5 file
+  output_path?: string;    // Optional custom output path
+}
+```
+
+### Output schema
+
+```typescript
+{
+  success: boolean;
+  ea: string;
+  inputs_found: number;
+  inputs: Array<{
+    name: string;
+    type: string;
+    default: string;
+    description: string | null;
+  }>;
+  set_file: string;  // Path to generated file
+  hint: string;
+}
+```
+
+---
+
+## `export_report`
+
+Export backtest report to various formats (CSV, JSON, Markdown) for external analysis or sharing.
+
+### Input schema
+
+```typescript
+{
+  report_dir: string;        // Path to backtest report directory
+  format?: "csv" | "json" | "md";  // Default: "csv"
+  output_path?: string;      // Optional custom output file path
+}
+```
+
+### Output schema
+
+```typescript
+{
+  success: boolean;
+  format: string;
+  output_file: string;
+  source: string;
+  hint: string;
+}
+```
+
+---
+
 ## `archive_report`

 Convert a backtest report directory into a compact JSON entry appended to `config/backtest_history.json`. Idempotent — re-archiving the same report is a no-op. Optionally deletes the source directory to reclaim disk space.
@@ -1527,6 +1966,31 @@ clean_cache(symbol=XAUUSD, dry_run=true)        → preview
 clean_cache(symbol=XAUUSD)                      → execute
 ```

+### Pre-flight validation
+
+```
+get_active_account()                            → current login, server, available symbols
+check_symbol_data_status(symbol=XAUUSD, from=2025.01.01, to=2025.03.31)
+                                                → verify data availability before backtest
+check_mt5_status()                              → verify MT5 installation and readiness
+validate_ea_syntax(path=MyEA.mq5)               → pre-compile syntax check
+```
+
+### Project management
+
+```
+init_project(name=MyStrategy, template=scalper)  → scaffold new EA with template
+create_set_template(ea=MyEA)                     → generate .set from EA inputs
+export_report(report_dir=..., format=csv)         → export to CSV/JSON/Markdown
+```
+
+### History and comparison
+
+```
+get_backtest_history(expert=MyEA, limit=10)       → list past backtests with metrics
+compare_backtests(report_dirs=["dir1", "dir2"])   → side-by-side comparison
+```
+
 ### Working with set files

 ```
@@ -6,19 +6,19 @@

 ```bash
 # macOS (Apple Silicon)
-curl -L -o mt5-quant.tar.gz https://github.com/masdevid/mt5-mcp/releases/latest/download/mt5-quant-macos-arm64.tar.gz
+curl -L -o mt5-quant.tar.gz https://github.com/masdevid/mt5-quant/releases/latest/download/mt5-quant-macos-arm64.tar.gz
 tar -xzf mt5-quant.tar.gz

 # Linux (x64)
-curl -L -o mt5-quant.tar.gz https://github.com/masdevid/mt5-mcp/releases/latest/download/mt5-quant-linux-x64.tar.gz
+curl -L -o mt5-quant.tar.gz https://github.com/masdevid/mt5-quant/releases/latest/download/mt5-quant-linux-x64.tar.gz
 tar -xzf mt5-quant.tar.gz
 ```

 ### Option B: Build from Source

 ```bash
-git clone https://github.com/masdevid/mt5-mcp
-cd mt5-mcp
+git clone https://github.com/masdevid/mt5-quant
+cd mt5-quant
 bash scripts/build-rust.sh
 ```

@@ -81,7 +81,7 @@ terminal_dir: "~/Library/Application Support/net.metaquotes.wine.metatrader5/dri
 ### Claude Code

 ```bash
-claude mcp add MT5-Quant -- /path/to/mt5-quant/target/release/mt5-quant
+claude mcp add io.github.masdevid/mt5-quant -- ~/.local/bin/mt5-quant
 ```

 Verify:
@@ -91,15 +91,70 @@ claude mcp list

 ### Windsurf

-Add to `~/.windsurf/config.yaml`:
-```yaml
-mcpServers:
-  mt5-quant:
-    command: /path/to/mt5-quant
-    env:
-      MT5_MCP_HOME: /path/to/mt5-mcp
+Add to `~/.codeium/windsurf/mcp_config.json`:
+
+```json
+{
+  "mcpServers": {
+    "io.github.masdevid/mt5-quant": {
+      "command": "~/.local/bin/mt5-quant",
+      "disabled": false,
+      "registry": "io.github.masdevid/mt5-quant"
+    }
+  }
+}
 ```

+### Cursor
+
+Add to `~/.cursor/mcp.json` (global) or `.cursor/mcp.json` (project):
+
+```json
+{
+  "mcpServers": {
+    "mt5-quant": {
+      "command": "~/.local/bin/mt5-quant"
+    }
+  }
+}
+```
+
+Or use Settings → MCP → Add Custom MCP.
+
+### VS Code
+
+Add to `.vscode/mcp.json` in your workspace:
+
+```json
+{
+  "servers": {
+    "mt5-quant": {
+      "command": "/path/to/mt5-quant"
+    }
+  }
+}
+```
+
+Or run `MCP: Add Server` from Command Palette.
+
+### Claude Desktop
+
+1. Open Agent panel → Click "..." menu → MCP Servers
+2. Click "Manage MCP Servers"
+3. Click "View raw config" or "Edit configuration"
+4. Add to `mcp_config.json`:
+
+```json
+{
+  "mcpServers": {
+    "mt5-quant": {
+      "command": "/path/to/mt5-quant"
+    }
+  }
+}
+```
+5. Restart Claude Desktop to apply changes
+
 ## 5. Verify Setup

 ```bash
@@ -0,0 +1,147 @@
+# VS Code MCP Integration Setup
+
+## Quick Setup
+
+### Option 1: Download Prebuilt Binary (Recommended)
+
+```bash
+# macOS (Apple Silicon)
+curl -L -o mt5-quant.tar.gz https://github.com/masdevid/mt5-quant/releases/latest/download/mt5-quant-macos-arm64.tar.gz
+tar -xzf mt5-quant.tar.gz
+
+# Linux (x64)
+curl -L -o mt5-quant.tar.gz https://github.com/masdevid/mt5-quant/releases/latest/download/mt5-quant-linux-x64.tar.gz
+tar -xzf mt5-quant.tar.gz
+```
+
+### Option 2: Build from Source
+
+```bash
+cargo build --release
+```
+
+## Configure VS Code
+
+### Method 1: Command Palette (Recommended)
+
+1. Press `Cmd/Ctrl + Shift + P`
+2. Run `MCP: Add Server`
+3. Choose **Workspace** or **User** scope
+4. Enter server name: `mt5-quant`
+5. Enter command: `~/.local/bin/mt5-quant`
+
+### Method 2: Edit mcp.json Directly
+
+Add to `.vscode/mcp.json` in your workspace:
+
+```json
+{
+  "servers": {
+    "mt5-quant": {
+      "command": "~/.local/bin/mt5-quant"
+    }
+  }
+}
+```
+
+Create the file:
+
+```bash
+mkdir -p .vscode
+cat > .vscode/mcp.json << 'EOF'
+{
+  "servers": {
+    "mt5-quant": {
+      "command": "~/.local/bin/mt5-quant"
+    }
+  }
+}
+EOF
+```
+
+### Method 3: VS Code CLI
+
+```bash
+code --add-mcp '{"name":"mt5-quant","command":"~/.local/bin/mt5-quant"}'
+```
+
+## Verify Setup
+
+In Copilot chat, type:
+
+```
+Run verify_setup
+```
+
+Expected output:
+```
+Wine:    /Applications/MetaTrader 5.app/.../wine64
+MT5 dir: ~/Library/Application Support/.../MetaTrader 5
+Display: gui
+Arch:    arch -x86_64
+```
+
+## Configuration Locations
+
+| Scope | Path | Use Case |
+|-------|------|----------|
+| Workspace | `.vscode/mcp.json` | Share with team via source control |
+| User | `~/.vscode/mcp.json` | Personal tools across all projects |
+| Dev Container | `devcontainer.json` → `customizations.vscode.mcp` | Containerized environments |
+
+## Troubleshooting
+
+### MCP server not appearing
+
+1. Open **Output** panel (`Cmd/Ctrl + Shift + U`)
+2. Select **MCP** from dropdown
+3. Check for connection errors
+4. Verify the path is absolute
+
+### Config not found
+
+The binary auto-detects its config, but you can also:
+1. Run `setup.sh` to create `config/mt5-quant.yaml`
+2. Or let the binary auto-discover on first run
+
+### Dev Container Setup
+
+Add to `.devcontainer/devcontainer.json`:
+
+```json
+{
+  "customizations": {
+    "vscode": {
+      "mcp": {
+        "servers": {
+          "mt5-quant": {
+            "command": "/path/to/mt5-quant"
+          }
+        }
+      }
+    }
+  }
+}
+```
+
+## Key Differences from Other IDEs
+
+VS Code uses `servers` (not `mcpServers`) in the JSON structure:
+
+```json
+{
+  "servers": {        // ← VS Code uses "servers"
+    "mt5-quant": {
+      "command": "..."
+    }
+  }
+}
+```
+
+Other platforms use `mcpServers`.
+
+## Resources
+
+- [VS Code MCP Documentation](https://code.visualstudio.com/docs/copilot/customization/mcp-servers)
+- [MCP Configuration Reference](https://code.visualstudio.com/docs/copilot/reference/mcp-configuration)
+- [MT5-Quant Tools Reference](./MCP_TOOLS.md)
@@ -6,11 +6,11 @@

 ```bash
 # macOS (Apple Silicon)
-curl -L -o mt5-quant.tar.gz https://github.com/masdevid/mt5-mcp/releases/latest/download/mt5-quant-macos-arm64.tar.gz
+curl -L -o mt5-quant.tar.gz https://github.com/masdevid/mt5-quant/releases/latest/download/mt5-quant-macos-arm64.tar.gz
 tar -xzf mt5-quant.tar.gz

 # Linux (x64)
-curl -L -o mt5-quant.tar.gz https://github.com/masdevid/mt5-mcp/releases/latest/download/mt5-quant-linux-x64.tar.gz
+curl -L -o mt5-quant.tar.gz https://github.com/masdevid/mt5-quant/releases/latest/download/mt5-quant-linux-x64.tar.gz
 tar -xzf mt5-quant.tar.gz
 ```

@@ -22,14 +22,28 @@ bash scripts/build-rust.sh

 ### 2. Configure Windsurf

-Edit `~/.windsurf/config.yaml`:
+Edit `~/.codeium/windsurf/mcp_config.json`:

-```yaml
-mcpServers:
-  mt5-quant:
-    command: /Users/masdevid/jobs/mt5-quant/target/release/mt5-quant
-    env:
-      MT5_MCP_HOME: /Users/masdevid/jobs/mt5-quant
+```json
+{
+  "mcpServers": {
+    "io.github.masdevid/mt5-quant": {
+      "command": "~/.local/bin/mt5-quant",
+      "disabled": false,
+      "registry": "io.github.masdevid/mt5-quant"
+    }
+  }
+}
+```
+
+Or use the automated setup:
+
+```bash
+# Install binary to standard location
+cp target/release/mt5-quant ~/.local/bin/
+
+# Then configure
+bash scripts/setup.sh
 ```

 ### 3. Restart Windsurf
@@ -66,12 +80,19 @@ scp -r config/mt5-quant.yaml user@server:~/.config/mt5-quant/config/
 - **NO Python required!**

 ### Windsurf Config on Target Machine
-```yaml
-mcpServers:
-  mt5-quant:
-    command: /usr/local/bin/mt5-quant
+
+```json
+{
+  "mcpServers": {
+    "mt5-quant": {
+      "command": "/usr/local/bin/mt5-quant"
+    }
+  }
+}
 ```

+The binary auto-detects its config location. No environment variables needed.
+
 ## Troubleshooting

 ### MCP server not appearing
@@ -1,8 +1,11 @@
-#!/bin/bash
-# Build release binaries for distribution
-# Creates: dist/mt5-quant-{platform}.tar.gz
+#!/usr/bin/env bash
+# Build release binary and package it for distribution
+# Creates: dist/mcp-mt5-quant-{platform}.tar.gz
+#
+# Usage:
+#   bash scripts/build-release.sh

-set -e
+set -euo pipefail

 SCRIPT_DIR="$(cd "$(dirname "${BASH_SOURCE[0]}")" && pwd)"
 PROJECT_ROOT="$(cd "${SCRIPT_DIR}/.." && pwd)"
@@ -10,16 +13,17 @@ PROJECT_ROOT="$(cd "${SCRIPT_DIR}/.." && pwd)"
 cd "$PROJECT_ROOT"

 VERSION=$(grep -E '^version = ' Cargo.toml | head -1 | sed 's/version = "\(.*\)"/\1/')
-echo "=== Building MCP-MT5-Quant v${VERSION} ==="
+echo "=== Building mt5-quant v${VERSION} ==="
 echo ""

 # Clean previous builds
 rm -rf "$PROJECT_ROOT/dist"
 mkdir -p "$PROJECT_ROOT/dist"

-# Build current platform
-echo "Building for current platform..."
+# Build release binary
+echo "Building release binary..."
 RUSTFLAGS="-D warnings" cargo build --release
+echo ""

 # Detect platform
 UNAME=$(uname -s)
@@ -30,43 +34,46 @@ if [[ "$UNAME" == "Darwin" ]]; then
 elif [[ "$UNAME" == "Linux" ]]; then
    PLATFORM="linux-${ARCH}"
 else
-    PLATFORM="unknown"
+    PLATFORM="unknown-${ARCH}"
 fi

 PACKAGE_NAME="mcp-mt5-quant-${PLATFORM}"
 PACKAGE_DIR="$PROJECT_ROOT/dist/${PACKAGE_NAME}"

 echo "Packaging for ${PLATFORM}..."
-mkdir -p "$PACKAGE_DIR"
+mkdir -p "$PACKAGE_DIR/docs"

-# Copy binary
+# Binary
 cp "$PROJECT_ROOT/target/release/mt5-quant" "$PACKAGE_DIR/"
+chmod +x "$PACKAGE_DIR/mt5-quant"

-# Copy config template
+# Config template
 mkdir -p "$PACKAGE_DIR/config"
 cp "$PROJECT_ROOT/config/mt5-quant.example.yaml" "$PACKAGE_DIR/config/"

-# Copy docs
-cp "$PROJECT_ROOT/README.md" "$PACKAGE_DIR/"
-cp "$PROJECT_ROOT/docs/WINDSURF.md" "$PACKAGE_DIR/WINDSURF_SETUP.md"
-cp "$PROJECT_ROOT/CLAUDE.md" "$PACKAGE_DIR/"
+# Documentation
+cp "$PROJECT_ROOT/README.md"            "$PACKAGE_DIR/"
+cp "$PROJECT_ROOT/docs/QUICKSTART.md"   "$PACKAGE_DIR/docs/"
+cp "$PROJECT_ROOT/docs/CLAUDE.md"       "$PACKAGE_DIR/docs/"
+cp "$PROJECT_ROOT/docs/CURSOR.md"       "$PACKAGE_DIR/docs/"
+cp "$PROJECT_ROOT/docs/VSCODE.md"       "$PACKAGE_DIR/docs/"
+cp "$PROJECT_ROOT/docs/WINDSURF.md"     "$PACKAGE_DIR/docs/"

 # Create tarball
 cd "$PROJECT_ROOT/dist"
 tar -czf "${PACKAGE_NAME}.tar.gz" "$PACKAGE_NAME"
+rm -rf "$PACKAGE_NAME"

 echo ""
-echo "=== Build Complete ==="
+echo "=== Build complete ==="
 echo ""
-echo "Package: dist/${PACKAGE_NAME}.tar.gz"
-echo "Binary: mt5-quant"
-echo "Size: $(du -h "${PACKAGE_NAME}.tar.gz" | cut -f1)"
+echo "Package : dist/${PACKAGE_NAME}.tar.gz"
+echo "Size    : $(du -h "${PACKAGE_NAME}.tar.gz" | cut -f1)"
 echo ""
 echo "Contents:"
-tar -tzf "${PACKAGE_NAME}.tar.gz" | head -10
+tar -tzf "${PACKAGE_NAME}.tar.gz"
 echo ""
-echo "To install:"
+echo "Install:"
 echo "  tar -xzf ${PACKAGE_NAME}.tar.gz"
-echo "  cd ${PACKAGE_NAME}"
-echo "  sudo cp mt5-quant /usr/local/bin/"
+echo "  sudo cp ${PACKAGE_NAME}/mt5-quant /usr/local/bin/"
 echo "  mt5-quant --help"
@@ -1,32 +1,25 @@
-#!/bin/bash
-# Build MT5-Quant Rust MCP Server
+#!/usr/bin/env bash
+# Build MT5-Quant release binary
 # Output: target/release/mt5-quant

-set -e
+set -euo pipefail

 SCRIPT_DIR="$(cd "$(dirname "${BASH_SOURCE[0]}")" && pwd)"
 PROJECT_ROOT="$(cd "${SCRIPT_DIR}/.." && pwd)"

 cd "$PROJECT_ROOT"

-echo "=== MT5-Quant Rust Build ==="
-echo "Project root: $PROJECT_ROOT"
+echo "=== MT5-Quant build ==="
+echo "Root: $PROJECT_ROOT"
 echo ""

-echo "Building release binary..."
 cargo build --release

 echo ""
-echo "=== Build Complete ==="
+echo "=== Done ==="
 echo ""
-echo "Executable location:"
 ls -lh "$PROJECT_ROOT/target/release/mt5-quant"
-
 echo ""
-echo "To test:"
-echo "  ./target/release/mt5-quant --help"
-echo ""
-echo "To install for Windsurf:"
-echo "  Update ~/.windsurf/config.yaml:"
-echo "    command: $PROJECT_ROOT/target/release/mt5-quant"
+echo "Run:     ./target/release/mt5-quant --help"
+echo "Install: cargo install --path . --force"
 echo ""
@@ -0,0 +1,238 @@
+#!/usr/bin/env bash
+# scripts/release.sh — Automate version bump, changelog, git tag, and push
+#
+# Usage:
+#   bash scripts/release.sh patch          # 1.31.0 → 1.31.1
+#   bash scripts/release.sh minor          # 1.31.0 → 1.32.0
+#   bash scripts/release.sh major          # 1.31.0 → 2.0.0
+#   bash scripts/release.sh 1.32.0         # explicit version
+#   bash scripts/release.sh v1.32.0        # with v prefix
+#   bash scripts/release.sh patch --yes    # non-interactive (CI / Claude Code)
+
+set -euo pipefail
+cd "$(dirname "$0")/.."
+
+RED='\033[0;31m'; GREEN='\033[0;32m'; YELLOW='\033[1;33m'; BLUE='\033[0;34m'; BOLD='\033[1m'; NC='\033[0m'
+info()  { echo -e "${BLUE}▶  $*${NC}"; }
+ok()    { echo -e "${GREEN}✓  $*${NC}"; }
+warn()  { echo -e "${YELLOW}⚠  $*${NC}"; }
+die()   { echo -e "${RED}✗  $*${NC}" >&2; exit 1; }
+hr()    { echo -e "${BLUE}────────────────────────────────────────${NC}"; }
+
+# ── Argument parsing ───────────────────────────────────────────────────────────
+
+bump="${1:-patch}"
+AUTO_YES=false
+if [[ "${2:-}" == "--yes" || "${2:-}" == "-y" || "${CI:-}" == "true" ]]; then
+  AUTO_YES=true
+fi
+
+# ── Prerequisites ──────────────────────────────────────────────────────────────
+
+command -v cargo   >/dev/null 2>&1 || die "cargo not found"
+command -v python3 >/dev/null 2>&1 || die "python3 not found"
+
+# ── Current version ────────────────────────────────────────────────────────────
+
+current=$(grep '^version' Cargo.toml | head -1 | sed 's/.*"\(.*\)".*/\1/')
+[[ -n "$current" ]] || die "Could not parse version from Cargo.toml"
+info "Current version: ${BOLD}$current${NC}"
+
+# ── Compute new version ────────────────────────────────────────────────────────
+
+IFS='.' read -r major minor patch_v <<< "$current"
+
+case "$bump" in
+  major)  new="$((major + 1)).0.0" ;;
+  minor)  new="${major}.$((minor + 1)).0" ;;
+  patch)  new="${major}.${minor}.$((patch_v + 1))" ;;
+  v*.*.*) new="${bump#v}" ;;
+  *.*.*)  new="$bump" ;;
+  *)      die "Usage: $0 [patch|minor|major|X.Y.Z] [--yes]" ;;
+esac
+
+[[ "$new" =~ ^[0-9]+\.[0-9]+\.[0-9]+$ ]] || die "Invalid version: $new"
+
+hr
+info "Releasing: ${BOLD}$current → $new${NC}"
+hr
+
+# ── Confirm ────────────────────────────────────────────────────────────────────
+
+if [[ "$AUTO_YES" == false ]]; then
+  read -rp "Proceed with release v${new}? [y/N] " confirm
+  [[ "$confirm" =~ ^[Yy]$ ]] || die "Aborted"
+else
+  info "Auto-confirmed (--yes)"
+fi
+
+# ── Check git state ────────────────────────────────────────────────────────────
+
+info "Checking git state..."
+if ! git diff --quiet 2>/dev/null || ! git diff --cached --quiet 2>/dev/null; then
+  die "Working tree has uncommitted changes — commit or stash first"
+fi
+current_branch=$(git rev-parse --abbrev-ref HEAD)
+info "Branch: $current_branch"
+
+if git rev-parse "v${new}" >/dev/null 2>&1; then
+  die "Tag v${new} already exists"
+fi
+ok "Git state clean"
+
+# ── 1. Bump Cargo.toml ─────────────────────────────────────────────────────────
+
+info "Bumping Cargo.toml..."
+if [[ "$(uname)" == "Darwin" ]]; then
+  sed -i '' "s/^version = \"${current}\"/version = \"${new}\"/" Cargo.toml
+else
+  sed -i "s/^version = \"${current}\"/version = \"${new}\"/" Cargo.toml
+fi
+cargo metadata --no-deps --format-version 1 >/dev/null 2>&1 || true
+ok "Cargo.toml → $new"
+
+# ── 2. Update server.json ──────────────────────────────────────────────────────
+
+info "Updating server.json..."
+NEW_VERSION="$new" python3 - <<'PYEOF'
+import json, os, re
+
+version = os.environ['NEW_VERSION']
+
+with open('server.json') as f:
+    data = json.load(f)
+
+data['version'] = version
+
+for pkg in data.get('packages', []):
+    pkg['version'] = version
+    if 'identifier' in pkg:
+        pkg['identifier'] = re.sub(
+            r'/v[0-9]+\.[0-9]+\.[0-9]+/',
+            f'/v{version}/',
+            pkg['identifier']
+        )
+    # SHA256 is computed by CI after building — placeholder signals this
+    pkg['fileSha256'] = 'TBD_CI_WILL_UPDATE'
+
+with open('server.json', 'w') as f:
+    json.dump(data, f, indent=2)
+    f.write('\n')
+
+print(f"  server.json version={version}, identifier URL updated")
+PYEOF
+ok "server.json → $new (SHA256 set by CI)"
+
+# ── 3. Generate CHANGELOG entry ────────────────────────────────────────────────
+
+info "Generating CHANGELOG entry..."
+prev_tag=$(git describe --tags --abbrev=0 2>/dev/null || echo "")
+today=$(date +%Y-%m-%d)
+
+{
+  echo "## [$new] — $today"
+  echo ""
+  if [[ -n "$prev_tag" ]]; then
+    git log "${prev_tag}..HEAD" --pretty=format:"- %s" --no-merges \
+      | grep -v "^- ci:" \
+      | grep -v "^- chore:" \
+      | grep -v "^$" \
+      || echo "- Minor improvements and bug fixes"
+  else
+    echo "- Initial public release"
+  fi
+  echo ""
+} > /tmp/release_entry.md
+
+if [[ -f CHANGELOG.md ]]; then
+  tmp=$(mktemp)
+  head -1 CHANGELOG.md > "$tmp"
+  echo "" >> "$tmp"
+  cat /tmp/release_entry.md >> "$tmp"
+  tail -n +2 CHANGELOG.md >> "$tmp"
+  mv "$tmp" CHANGELOG.md
+else
+  { echo "# Changelog"; echo ""; cat /tmp/release_entry.md; } > CHANGELOG.md
+fi
+rm -f /tmp/release_entry.md
+ok "CHANGELOG.md updated"
+
+# ── 4. Verify build ────────────────────────────────────────────────────────────
+
+info "Verifying build (cargo check)..."
+cargo check --quiet 2>&1 || die "cargo check failed — fix errors before releasing"
+ok "Build check passed"
+
+# ── 5. Commit ──────────────────────────────────────────────────────────────────
+
+info "Creating release commit..."
+git add Cargo.toml Cargo.lock server.json CHANGELOG.md
+git commit -m "release: v${new}
+
+- Version bump ${current} → ${new}
+- server.json identifier URL updated (SHA256 set by CI after build)
+- CHANGELOG.md updated"
+ok "Release commit created"
+
+# ── 6. Tag ─────────────────────────────────────────────────────────────────────
+
+info "Creating annotated tag v${new}..."
+git tag -a "v${new}" -m "Release v${new}"
+ok "Tagged v${new}"
+
+# ── 7. Push (rebase if remote has new commits from CI) ─────────────────────────
+
+info "Pushing to GitHub..."
+if ! git push origin "$current_branch" 2>/dev/null; then
+  warn "Push rejected — rebasing on remote (CI may have committed SHA256)..."
+  git fetch origin "$current_branch"
+
+  # Resolve server.json conflict automatically: keep our version
+  if ! git rebase "origin/${current_branch}"; then
+    if git diff --name-only --diff-filter=U | grep -q "server.json"; then
+      python3 - <<'PYEOF'
+import re
+with open('server.json') as f:
+    raw = f.read()
+resolved = re.sub(
+    r'<<<<<<< HEAD.*?=======\n(.*?)>>>>>>> [^\n]+\n',
+    r'\1',
+    raw,
+    flags=re.DOTALL
+)
+with open('server.json', 'w') as f:
+    f.write(resolved)
+PYEOF
+      git add server.json
+      GIT_EDITOR=true git rebase --continue
+    else
+      git rebase --abort
+      die "Rebase failed with unexpected conflicts — push manually"
+    fi
+  fi
+
+  git push origin "$current_branch"
+fi
+
+# Push tag (delete and re-push if rebase moved the commit)
+git push origin "v${new}" 2>/dev/null || {
+  git push origin ":refs/tags/v${new}" 2>/dev/null || true
+  git tag -f -a "v${new}" -m "Release v${new}"
+  git push origin "v${new}"
+}
+
+ok "Pushed — GitHub Actions triggered"
+
+# ── Done ───────────────────────────────────────────────────────────────────────
+
+hr
+echo ""
+echo -e "${GREEN}${BOLD}  Release v${new} kicked off!${NC}"
+echo ""
+echo -e "  Actions:  https://github.com/masdevid/mt5-quant/actions"
+echo -e "  Release:  https://github.com/masdevid/mt5-quant/releases/tag/v${new}"
+echo ""
+echo -e "  ${YELLOW}CI will compute the MCP package SHA256 and${NC}"
+echo -e "  ${YELLOW}commit it back to server.json automatically.${NC}"
+echo ""
+hr
@@ -7,6 +7,16 @@
 # --keep-last N    Keep only last N backtest reports (default: 20)
 # --claude-code    Generate CLAUDE.md template and .claude/hooks/user-prompt-submit.sh
 #                  (skips main config wizard — run standalone or alongside normal setup)
+#
+# MCP Auto-Registration (per official 2025 docs):
+#   Automatically detects and registers with available MCP platforms:
+#   - Claude Code    : via 'claude mcp add' (stored in ~/.claude.json)
+#   - Windsurf     : ~/.codeium/windsurf/mcp_config.json (JSON, mcpServers)
+#   - Cursor       : ~/.cursor/mcp.json (JSON, mcpServers)
+#   - VS Code      : .vscode/mcp.json (JSON, servers - not mcpServers)
+#   - Claude Desktop: ~/Library/Application Support/Claude/claude_desktop_config.json
+#
+#   Previous installations are auto-detected and uninstalled before re-registering.

 set -uo pipefail

@@ -648,6 +658,30 @@ main() {
    _bold "MT5-Quant setup — auto-detecting Wine and MT5 paths"
    echo  "────────────────────────────────────────────────────"

+    # ── Check for previous installation ────────────────────────────────────────
+    if _check_any_existing_registration; then
+        _yellow "Previous mt5-quant MCP installation detected on one or more platforms"
+        echo "  New path: ${REPO_DIR}/server/main.py"
+
+        local reinstall_ans="yes"
+        if ! $AUTO_YES; then
+            reinstall_ans=$(_ask "Unregister all previous installations and reinstall?" "yes")
+        fi
+
+        if [[ "$reinstall_ans" =~ ^[Yy] ]]; then
+            _unregister_all_platforms || {
+                if ! $AUTO_YES; then
+                    local force_ans
+                    force_ans=$(_ask "Some platforms failed to unregister. Continue anyway?" "no")
+                    [[ ! "$force_ans" =~ ^[Yy] ]] && exit 1
+                fi
+            }
+        else
+            echo "Aborted — keeping existing installations."
+            exit 0
+        fi
+    fi
+
    # ── Check if config already exists ───────────────────────────────────────
    if [[ -f "$CONFIG_OUT" ]] && ! $AUTO_YES; then
        _yellow "Config already exists: $CONFIG_OUT"
@@ -784,49 +818,454 @@ main() {
    _ok "Written: $CONFIG_OUT"
    echo "  Tip: see config/example.set for optimization .set file format"

-    # ── Register with Claude Code ──────────────────────────────────────────────
-    _offer_mcp_register
+    # ── Register with all detected MCP platforms ─────────────────────────────
+    _register_all_mcp_platforms

    echo ""
    _green "Setup complete!"
    echo ""
 }

-# ── MCP registration ──────────────────────────────────────────────────────────
-_offer_mcp_register() {
-    echo ""
-    _bold "Registering with Claude Code..."
+# ── MCP Platform Detection & Registration ───────────────────────────────────

-    if ! command -v claude &>/dev/null; then
-        _warn "claude CLI not found — register manually:"
+# Detect available MCP platforms and return them as a list
+detect_mcp_platforms() {
+    local platforms=()
+
+    # Claude Code
+    if command -v claude &>/dev/null; then
+        platforms+=("claude")
+    fi
+
+    # Windsurf (uses ~/.codeium/windsurf/mcp_config.json)
+    if [[ -d "$HOME/.codeium/windsurf" ]] || [[ -d "$HOME/.windsurf" ]] || command -v windsurf &>/dev/null; then
+        platforms+=("windsurf")
+    fi
+
+    # Cursor (uses ~/.cursor/mcp.json)
+    if [[ -d "$HOME/.cursor" ]] || [[ -d "$HOME/Library/Application Support/Cursor" ]] || command -v cursor &>/dev/null; then
+        platforms+=("cursor")
+    fi
+
+    # VS Code (uses .vscode/mcp.json or user profile)
+    if [[ -d "$HOME/.vscode" ]] || [[ -d "$HOME/Library/Application Support/Code" ]] || command -v code &>/dev/null; then
+        platforms+=("vscode")
+    fi
+
+    printf '%s\n' "${platforms[@]:-}"
+}
+
+# Check if mt5-quant is registered on a specific platform
+_is_registered_on_platform() {
+    local platform="$1"
+    case "$platform" in
+        claude)
+            if command -v claude &>/dev/null; then
+                local mcp_list
+                mcp_list=$(claude mcp list 2>/dev/null || true)
+                echo "$mcp_list" | grep -q "mt5-quant"
+                return $?
+            fi
+            return 1
+            ;;
+        windsurf)
+            local config_file="$HOME/.codeium/windsurf/mcp_config.json"
+            [[ -f "$config_file" ]] && grep -q '"mt5-quant"' "$config_file" 2>/dev/null
+            return $?
+            ;;
+        cursor)
+            local config_file
+            config_file="$HOME/.cursor/mcp.json"
+            [[ -f "$config_file" ]] && grep -q "mt5-quant" "$config_file" 2>/dev/null
+            return $?
+            ;;
+        vscode)
+            # VS Code uses .vscode/mcp.json in workspace or user profile
+            local workspace_config=".vscode/mcp.json"
+            local user_config
+            user_config="$HOME/.vscode/mcp.json"
+            [[ -f "$workspace_config" ]] && grep -q '"mt5-quant"' "$workspace_config" 2>/dev/null && return 0
+            [[ -f "$user_config" ]] && grep -q '"mt5-quant"' "$user_config" 2>/dev/null && return 0
+            return 1
+            ;;
+    esac
+    return 1
+}
+
+# Get registered path for a platform
+_get_platform_mcp_path() {
+    local platform="$1"
+    case "$platform" in
+        claude)
+            if command -v claude &>/dev/null; then
+                local mcp_list
+                mcp_list=$(claude mcp list 2>/dev/null || true)
+                echo "$mcp_list" | grep "mt5-quant" | head -1 | sed -E 's/.*--[[:space:]]*//' | tr -d ' '
+            fi
+            ;;
+        windsurf)
+            local config_file="$HOME/.codeium/windsurf/mcp_config.json"
+            if [[ -f "$config_file" ]]; then
+                grep -A3 '"mt5-quant"' "$config_file" 2>/dev/null | grep '"command"' | sed 's/.*"command":[[:space:]]*"\([^"]*\)".*/\1/'
+            fi
+            ;;
+        cursor)
+            local config_file="$HOME/.cursor/mcp.json"
+            if [[ -f "$config_file" ]]; then
+                grep -A3 '"mt5-quant"' "$config_file" 2>/dev/null | grep '"command"' | sed 's/.*"command":[[:space:]]*"\([^"]*\)".*/\1/'
+            fi
+            ;;
+    esac
+}
+
+# Unregister from a specific platform
+_unregister_from_platform() {
+    local platform="$1"
+    echo ""
+    _bold "Unregistering from $platform..."
+
+    case "$platform" in
+        claude)
+            if ! command -v claude &>/dev/null; then
+                _warn "claude CLI not found"
+                return 1
+            fi
+            local out
+            out=$(claude mcp remove mt5-quant 2>&1) || true
+            if echo "$out" | grep -qi "removed\|success\|deleted"; then
+                _ok "Unregistered from Claude Code"
+                return 0
+            elif echo "$out" | grep -qi "not found\|does not exist"; then
+                _ok "No existing registration on Claude Code"
+                return 0
+            else
+                _warn "Unregister result: $out"
+                return 1
+            fi
+            ;;
+        windsurf)
+            local config_file="$HOME/.codeium/windsurf/mcp_config.json"
+            if [[ -f "$config_file" ]]; then
+                # Remove mt5-quant from JSON using Python
+                if command -v python3 &>/dev/null; then
+                    python3 -c "
+import json, sys
+with open('$config_file') as f:
+    data = json.load(f)
+if 'mcpServers' in data and 'mt5-quant' in data['mcpServers']:
+    del data['mcpServers']['mt5-quant']
+    with open('$config_file', 'w') as f:
+        json.dump(data, f, indent=2)
+" && { _ok "Unregistered from Windsurf"; return 0; }
+                fi
+                _warn "Could not auto-unregister from Windsurf — edit $config_file manually"
+                return 1
+            fi
+            _ok "No existing registration on Windsurf"
+            return 0
+            ;;
+        cursor)
+            local config_file="$HOME/.cursor/mcp.json"
+            if [[ -f "$config_file" ]]; then
+                # Remove mt5-quant from JSON using Python if available, or sed as fallback
+                if command -v python3 &>/dev/null; then
+                    python3 -c "
+import json, sys
+with open('$config_file') as f:
+    data = json.load(f)
+if 'mcpServers' in data and 'mt5-quant' in data['mcpServers']:
+    del data['mcpServers']['mt5-quant']
+    with open('$config_file', 'w') as f:
+        json.dump(data, f, indent=2)
+" && { _ok "Unregistered from Cursor"; return 0; }
+                fi
+                _warn "Could not auto-unregister from Cursor — edit $config_file manually"
+                return 1
+            fi
+            _ok "No existing registration on Cursor"
+            return 0
+            ;;
+        vscode)
+            # VS Code uses 'servers' (not 'mcpServers') in mcp.json
+            local workspace_config=".vscode/mcp.json"
+            local user_config="$HOME/.vscode/mcp.json"
+            local config_file=""
+            
+            # Determine which config file to use
+            if [[ -f "$workspace_config" ]] && grep -q '"mt5-quant"' "$workspace_config" 2>/dev/null; then
+                config_file="$workspace_config"
+            elif [[ -f "$user_config" ]] && grep -q '"mt5-quant"' "$user_config" 2>/dev/null; then
+                config_file="$user_config"
+            fi
+            
+            if [[ -n "$config_file" ]]; then
+                if command -v python3 &>/dev/null; then
+                    python3 -c "
+import json, sys
+with open('$config_file') as f:
+    data = json.load(f)
+if 'servers' in data and 'mt5-quant' in data['servers']:
+    del data['servers']['mt5-quant']
+    with open('$config_file', 'w') as f:
+        json.dump(data, f, indent=2)
+" && { _ok "Unregistered from VS Code ($config_file)"; return 0; }
+                fi
+                _warn "Could not auto-unregister from VS Code — edit $config_file manually"
+                return 1
+            fi
+            _ok "No existing registration on VS Code"
+            return 0
+            ;;
+    esac
+}
+
+# Register with a specific platform
+_register_with_platform() {
+    local platform="$1"
+    local use_binary="${2:-false}"
+
+    # Determine command path - use standard installation location
+    local cmd_path
+    local default_install="$HOME/.local/bin/mt5-quant"
+    
+    if [[ -f "$default_install" ]]; then
+        cmd_path="$default_install"
+    elif [[ -f "${REPO_DIR}/target/release/mt5-quant" ]]; then
+        # Fallback to project build if standard location not found
+        cmd_path="${REPO_DIR}/target/release/mt5-quant"
+        _warn "Using project build at $cmd_path"
+        _warn "Consider installing to standard location: cp $cmd_path ~/.local/bin/"
+    else
+        _fail "mt5-quant not found at $default_install or ${REPO_DIR}/target/release/"
+        return 1
+    fi
+
+    case "$platform" in
+        claude)
+            local out
+            out=$(claude mcp add mt5-quant -- $cmd_path 2>&1) || true
+            if echo "$out" | grep -qi "already\|exists"; then
+                _ok "Already registered on Claude Code"
+            elif echo "$out" | grep -qi "error\|failed"; then
+                _warn "Claude Code registration failed: $out"
+                return 1
+            else
+                _ok "Registered on Claude Code"
+            fi
+            ;;
+        windsurf)
+            local config_file="$HOME/.codeium/windsurf/mcp_config.json"
+            mkdir -p "$(dirname "$config_file")"
+
+            # Create or update JSON config
+            if command -v python3 &>/dev/null; then
+                python3 -c "
+import json, os
+config_path = '$config_file'
+data = {'mcpServers': {}}
+if os.path.exists(config_path):
+    try:
+        with open(config_path) as f:
+            data = json.load(f)
+    except:
+        pass
+if 'mcpServers' not in data:
+    data['mcpServers'] = {}
+
+data['mcpServers']['io.github.masdevid/mt5-quant'] = {
+    'command': '$cmd_path',
+    'disabled': False,
+    'registry': 'io.github.masdevid/mt5-quant'
+}
+
+with open(config_path, 'w') as f:
+    json.dump(data, f, indent=2)
+" && { _ok "Registered on Windsurf ($config_file)"; return 0; }
+            else
+                _warn "Python3 required for Windsurf registration"
+                return 1
+            fi
+            ;;
+        cursor)
+            local config_file="$HOME/.cursor/mcp.json"
+            mkdir -p "$(dirname "$config_file")"
+
+            # Create or update JSON config
+            if command -v python3 &>/dev/null; then
+                python3 -c "
+import json, os
+config_path = '$config_file'
+data = {'mcpServers': {}}
+if os.path.exists(config_path):
+    try:
+        with open(config_path) as f:
+            data = json.load(f)
+    except:
+        pass
+if 'mcpServers' not in data:
+    data['mcpServers'] = {}
+
+data['mcpServers']['mt5-quant'] = {
+    'command': '$cmd_path'
+}
+
+with open(config_path, 'w') as f:
+    json.dump(data, f, indent=2)
+" && { _ok "Registered on Cursor ($config_file)"; return 0; }
+            else
+                _warn "Python3 required for Cursor registration"
+                return 1
+            fi
+            ;;
+        vscode)
+            # VS Code uses .vscode/mcp.json (workspace) or user profile
+            # Format: { "servers": { "name": { "command": "...", "args": [...], "env": {...} } }
+            local workspace_config=".vscode/mcp.json"
+            mkdir -p ".vscode"
+            
+            if command -v python3 &>/dev/null; then
+                python3 -c "
+import json, os
+config_path = '$workspace_config'
+data = {'servers': {}}
+if os.path.exists(config_path):
+    try:
+        with open(config_path) as f:
+            data = json.load(f)
+    except:
+        pass
+if 'servers' not in data:
+    data['servers'] = {}
+
+# For stdio servers, VS Code uses 'command' and optional 'args'
+data['servers']['mt5-quant'] = {
+    'command': '$cmd_path'
+}
+
+with open(config_path, 'w') as f:
+    json.dump(data, f, indent=2)
+" && { _ok "Registered on VS Code ($workspace_config)"; return 0; }
+            else
+                _warn "Python3 required for VS Code registration"
+                return 1
+            fi
+            ;;
+    esac
+}
+
+# Check for any existing registrations across all platforms
+_check_any_existing_registration() {
+    local platforms=()
+    while IFS= read -r platform; do
+        [[ -n "$platform" ]] && platforms+=("$platform")
+    done < <(detect_mcp_platforms)
+
+    local found=false
+    for platform in "${platforms[@]}"; do
+        if _is_registered_on_platform "$platform"; then
+            found=true
+            break
+        fi
+    done
+
+    $found
+}
+
+# Unregister from all platforms
+_unregister_all_platforms() {
+    local platforms=()
+    while IFS= read -r platform; do
+        [[ -n "$platform" ]] && platforms+=("$platform")
+    done < <(detect_mcp_platforms)
+
+    for platform in "${platforms[@]}"; do
+        _is_registered_on_platform "$platform" && _unregister_from_platform "$platform"
+    done
+}
+
+# Main registration function - detects platforms and registers with all
+_register_all_mcp_platforms() {
+    echo ""
+    _bold "Detecting MCP platforms..."
+
+    local platforms=()
+    while IFS= read -r platform; do
+        [[ -n "$platform" ]] && platforms+=("$platform")
+    done < <(detect_mcp_platforms)
+
+    if [[ ${#platforms[@]} -eq 0 ]]; then
+        _warn "No MCP platforms detected (Claude, Windsurf, Cursor, VS Code)"
        echo ""
-        echo "  claude mcp add mt5-quant -- python3 \"${REPO_DIR}/server/main.py\""
+        echo "  Manual registration required:"
+        echo "  - Claude Code: claude mcp add mt5-quant -- python3 \"${REPO_DIR}/server/main.py\""
+        echo "  - Windsurf:   Edit ~/.codeium/windsurf/mcp_config.json"
+        echo "  - Cursor:     Edit ~/.cursor/mcp.json"
+        echo "  - VS Code:    Edit .vscode/mcp.json (workspace) or use MCP: Add Server command"
+        echo "  - Claude Desktop: Edit ~/Library/Application Support/Claude/claude_desktop_config.json"
        echo ""
        return
    fi

-    local register=true
-    if ! $AUTO_YES; then
-        local ans
-        ans=$(_ask "Register mt5-quant with Claude Code now?" "yes")
-        [[ ! "$ans" =~ ^[Yy] ]] && register=false
+    _ok "Found platforms: ${platforms[*]}"
+
+    # Check for existing registrations
+    local has_existing=false
+    for platform in "${platforms[@]}"; do
+        if _is_registered_on_platform "$platform"; then
+            has_existing=true
+            local old_path
+            old_path=$(_get_platform_mcp_path "$platform")
+            _yellow "Existing registration detected on $platform"
+            [[ -n "$old_path" ]] && echo "  Current path: $old_path"
+        fi
+    done
+
+    # Prompt for reinstall if any existing registrations found
+    if $has_existing && ! $AUTO_YES; then
+        echo ""
+        local reinstall_ans
+        reinstall_ans=$(_ask "Unregister existing installations and reinstall on all platforms?" "yes")
+        if [[ ! "$reinstall_ans" =~ ^[Yy] ]]; then
+            echo "  Keeping existing registrations."
+            return
+        fi
    fi

-    if $register; then
-        local out
-        out=$(claude mcp add mt5-quant -- python3 "${REPO_DIR}/server/main.py" 2>&1) || true
-        if echo "$out" | grep -qi "already\|exists"; then
-            _ok "Already registered (no change needed)"
-        elif echo "$out" | grep -qi "error\|failed"; then
-            _warn "Registration failed: $out"
-            echo "  Run manually: claude mcp add mt5-quant -- python3 \"${REPO_DIR}/server/main.py\""
-        else
-            _ok "Registered: claude mcp add mt5-quant"
-        fi
-    else
-        echo "  Skipped. Run manually:"
-        echo "  claude mcp add mt5-quant -- python3 \"${REPO_DIR}/server/main.py\""
-    fi
+    # Unregister from all platforms first
+    for platform in "${platforms[@]}"; do
+        _is_registered_on_platform "$platform" && _unregister_from_platform "$platform"
+    done
+
+    # Register with all detected platforms
+    echo ""
+    _bold "Registering mt5-quant MCP..."
+
+    for platform in "${platforms[@]}"; do
+        # Use binary for Windsurf/Cursor, Python for Claude
+        case "$platform" in
+            windsurf|cursor)
+                _register_with_platform "$platform" true
+                ;;
+            *)
+                _register_with_platform "$platform" false
+                ;;
+        esac
+    done
+
+    echo ""
+    _green "MCP registration complete!"
+    echo ""
+    echo "  Registered on: ${platforms[*]}"
+    echo ""
+    echo "  Config locations:"
+    [[ " ${platforms[*]} " =~ " claude " ]] && echo "    Claude Code:   ~/.claude.json (managed via CLI)"
+    [[ " ${platforms[*]} " =~ " windsurf " ]] && echo "    Windsurf:      ~/.codeium/windsurf/mcp_config.json"
+    [[ " ${platforms[*]} " =~ " cursor " ]] && echo "    Cursor:        ~/.cursor/mcp.json"
+    [[ " ${platforms[*]} " =~ " vscode " ]] && echo "    VS Code:       .vscode/mcp.json"
+    echo ""
+    echo "  Binary: ${REPO_DIR}/target/release/mt5-quant"
+    echo "  Python: ${REPO_DIR}/server/main.py"
+    echo ""
 }

 main
@@ -1,18 +1,19 @@
 {
-  "$schema": "https://raw.githubusercontent.com/modelcontextprotocol/specification/main/schema/2024-11-05/schema.json",
+  "$schema": "https://static.modelcontextprotocol.io/schemas/2025-12-11/server.schema.json",
  "name": "io.github.masdevid/mt5-quant",
  "title": "MT5-Quant",
-  "description": "MCP server for MT5 strategy development. Compile, backtest, analyze MQL5 EAs on macOS/Linux.",
+  "description": "MT5 strategy development with 87 tools. Backtest, optimize, debug MQL5 EAs on macOS/Linux.",
  "repository": {
    "url": "https://github.com/masdevid/mt5-quant",
    "source": "github"
  },
-  "version": "1.27.0",
+  "version": "1.31.5",
  "packages": [
    {
      "registryType": "mcpb",
-      "identifier": "https://github.com/masdevid/mt5-quant/releases/download/v1.27.0/mcp-mt5-quant-macos-arm64.tar.gz",
-      "fileSha256": "e4dd26f2fa920e22575c88546e4d0dc54ec34d51950f3c58bfdf3a3f10b8a9a4",
+      "version": "1.31.5",
+      "identifier": "https://github.com/masdevid/mt5-quant/releases/download/v1.31.5/mcp-mt5-quant-macos-arm64.tar.gz",
+      "fileSha256": "TBD_CI_WILL_UPDATE",
      "transport": {
        "type": "stdio"
      },
@@ -1,4 +1,4 @@
-use chrono::{DateTime, NaiveDateTime};
+use chrono::{DateTime, Datelike, NaiveDateTime, Timelike};
 use serde::{Deserialize, Serialize};
 use std::collections::HashMap;

@@ -489,3 +489,612 @@ pub struct ConcurrentPeak {
    pub peak_open: i32,
    pub peak_time: String,
 }
+
+#[derive(Debug, Clone, Serialize, Deserialize)]
+pub struct ProfitDistribution {
+    pub small_wins: i32,
+    pub medium_wins: i32,
+    pub large_wins: i32,
+    pub small_losses: i32,
+    pub medium_losses: i32,
+    pub large_losses: i32,
+    pub small_win_pnl: f64,
+    pub medium_win_pnl: f64,
+    pub large_win_pnl: f64,
+    pub small_loss_pnl: f64,
+    pub medium_loss_pnl: f64,
+    pub large_loss_pnl: f64,
+    pub buckets: Vec<ProfitBucket>,
+}
+
+#[derive(Debug, Clone, Serialize, Deserialize)]
+pub struct ProfitBucket {
+    pub range: String,
+    pub min: f64,
+    pub max: f64,
+    pub count: i32,
+    pub total_pnl: f64,
+}
+
+#[derive(Debug, Clone, Serialize, Deserialize)]
+pub struct TimePerformance {
+    pub by_hour: Vec<HourPerformance>,
+    pub by_day: Vec<DayPerformance>,
+    pub best_hour: i32,
+    pub worst_hour: i32,
+    pub best_day: String,
+    pub worst_day: String,
+}
+
+#[derive(Debug, Clone, Serialize, Deserialize)]
+pub struct HourPerformance {
+    pub hour: i32,
+    pub trades: i32,
+    pub wins: i32,
+    pub total_pnl: f64,
+    pub win_rate: f64,
+}
+
+#[derive(Debug, Clone, Serialize, Deserialize)]
+pub struct DayPerformance {
+    pub day: String,
+    pub day_num: i32,
+    pub trades: i32,
+    pub wins: i32,
+    pub total_pnl: f64,
+    pub win_rate: f64,
+}
+
+#[derive(Debug, Clone, Serialize, Deserialize)]
+pub struct HoldTimeAnalysis {
+    pub avg_hold_minutes: f64,
+    pub median_hold_minutes: f64,
+    pub buckets: Vec<HoldTimeBucket>,
+    pub correlation_with_profit: f64,
+}
+
+#[derive(Debug, Clone, Serialize, Deserialize)]
+pub struct HoldTimeBucket {
+    pub range: String,
+    pub min_minutes: f64,
+    pub max_minutes: f64,
+    pub count: i32,
+    pub avg_profit: f64,
+    pub total_pnl: f64,
+}
+
+#[derive(Debug, Clone, Serialize, Deserialize)]
+pub struct LayerPerformance {
+    pub layer: i32,
+    pub trades: i32,
+    pub wins: i32,
+    pub total_pnl: f64,
+    pub win_rate: f64,
+    pub avg_volume: f64,
+    pub avg_profit: f64,
+}
+
+#[derive(Debug, Clone, Serialize, Deserialize)]
+pub struct VolumeAnalysis {
+    pub correlation_with_profit: f64,
+    pub by_volume_bucket: Vec<VolumeBucket>,
+}
+
+#[derive(Debug, Clone, Serialize, Deserialize)]
+pub struct VolumeBucket {
+    pub volume_range: String,
+    pub min_volume: f64,
+    pub max_volume: f64,
+    pub trades: i32,
+    pub avg_profit: f64,
+    pub total_pnl: f64,
+}
+
+#[derive(Debug, Clone, Serialize, Deserialize)]
+pub struct CostAnalysis {
+    pub total_commission: f64,
+    pub total_swap: f64,
+    pub commission_pct_of_profit: f64,
+    pub swap_pct_of_profit: f64,
+    pub avg_commission_per_trade: f64,
+    pub avg_swap_per_trade: f64,
+    pub net_profit_before_costs: f64,
+    pub cost_impact_on_win_rate: f64,
+}
+
+#[derive(Debug, Clone, Serialize, Deserialize)]
+pub struct EfficiencyAnalysis {
+    pub profit_per_hour: f64,
+    pub profit_per_day: f64,
+    pub profit_per_trade_hour: f64,
+    pub avg_trade_duration_hours: f64,
+    pub annualized_return_pct: f64,
+    pub trades_per_day: f64,
+}
+
+impl DealAnalyzer {
+    pub fn profit_distribution(&self, deals: &[Deal]) -> ProfitDistribution {
+        let closed: Vec<&Deal> = deals
+            .iter()
+            .filter(|d| d.entry.to_lowercase().contains("out") && d.profit != 0.0)
+            .collect();
+
+        let mut small_wins = 0;
+        let mut medium_wins = 0;
+        let mut large_wins = 0;
+        let mut small_losses = 0;
+        let mut medium_losses = 0;
+        let mut large_losses = 0;
+        let mut small_win_pnl = 0.0;
+        let mut medium_win_pnl = 0.0;
+        let mut large_win_pnl = 0.0;
+        let mut small_loss_pnl = 0.0;
+        let mut medium_loss_pnl = 0.0;
+        let mut large_loss_pnl = 0.0;
+
+        for deal in &closed {
+            let profit = deal.profit;
+            if profit > 0.0 {
+                if profit < 50.0 {
+                    small_wins += 1;
+                    small_win_pnl += profit;
+                } else if profit < 200.0 {
+                    medium_wins += 1;
+                    medium_win_pnl += profit;
+                } else {
+                    large_wins += 1;
+                    large_win_pnl += profit;
+                }
+            } else {
+                let loss = profit.abs();
+                if loss < 50.0 {
+                    small_losses += 1;
+                    small_loss_pnl += profit;
+                } else if loss < 200.0 {
+                    medium_losses += 1;
+                    medium_loss_pnl += profit;
+                } else {
+                    large_losses += 1;
+                    large_loss_pnl += profit;
+                }
+            }
+        }
+
+        // Create detailed buckets
+        let bucket_ranges = [
+            (-999999.0, -500.0, "Loss $500+"),
+            (-500.0, -200.0, "Loss $200-500"),
+            (-200.0, -50.0, "Loss $50-200"),
+            (-50.0, 0.0, "Loss $0-50"),
+            (0.0, 50.0, "Win $0-50"),
+            (50.0, 200.0, "Win $50-200"),
+            (200.0, 500.0, "Win $200-500"),
+            (500.0, 999999.0, "Win $500+"),
+        ];
+
+        let mut buckets: Vec<ProfitBucket> = bucket_ranges
+            .iter()
+            .map(|(min, max, range)| {
+                let count = closed
+                    .iter()
+                    .filter(|d| d.profit >= *min && d.profit < *max)
+                    .count() as i32;
+                let total_pnl: f64 = closed
+                    .iter()
+                    .filter(|d| d.profit >= *min && d.profit < *max)
+                    .map(|d| d.profit)
+                    .sum();
+                ProfitBucket {
+                    range: range.to_string(),
+                    min: *min,
+                    max: *max,
+                    count,
+                    total_pnl: (total_pnl * 100.0).round() / 100.0,
+                }
+            })
+            .collect();
+
+        // Remove empty buckets
+        buckets.retain(|b| b.count > 0);
+
+        ProfitDistribution {
+            small_wins,
+            medium_wins,
+            large_wins,
+            small_losses,
+            medium_losses,
+            large_losses,
+            small_win_pnl: (small_win_pnl * 100.0).round() / 100.0,
+            medium_win_pnl: (medium_win_pnl * 100.0).round() / 100.0,
+            large_win_pnl: (large_win_pnl * 100.0).round() / 100.0,
+            small_loss_pnl: (small_loss_pnl * 100.0).round() / 100.0,
+            medium_loss_pnl: (medium_loss_pnl * 100.0).round() / 100.0,
+            large_loss_pnl: (large_loss_pnl * 100.0).round() / 100.0,
+            buckets,
+        }
+    }
+
+    pub fn time_performance(&self, deals: &[Deal]) -> TimePerformance {
+        let closed: Vec<&Deal> = deals
+            .iter()
+            .filter(|d| d.entry.to_lowercase().contains("out") && d.profit != 0.0)
+            .collect();
+
+        let mut hourly: HashMap<i32, (i32, i32, f64)> = HashMap::new();
+        let mut daily: HashMap<String, (i32, i32, f64, i32)> = HashMap::new();
+
+        for deal in &closed {
+            if let Some(dt) = Self::parse_datetime(&deal.time) {
+                let hour = dt.hour() as i32;
+                let day_num = dt.weekday().num_days_from_monday() as i32;
+                let day_name = match dt.weekday() {
+                    chrono::Weekday::Mon => "Mon",
+                    chrono::Weekday::Tue => "Tue",
+                    chrono::Weekday::Wed => "Wed",
+                    chrono::Weekday::Thu => "Thu",
+                    chrono::Weekday::Fri => "Fri",
+                    chrono::Weekday::Sat => "Sat",
+                    chrono::Weekday::Sun => "Sun",
+                }.to_string();
+
+                let entry = hourly.entry(hour).or_insert((0, 0, 0.0));
+                entry.0 += 1;
+                entry.2 += deal.profit;
+                if deal.profit > 0.0 {
+                    entry.1 += 1;
+                }
+
+                let day_entry = daily.entry(day_name.clone()).or_insert((0, 0, 0.0, day_num));
+                day_entry.0 += 1;
+                day_entry.2 += deal.profit;
+                if deal.profit > 0.0 {
+                    day_entry.1 += 1;
+                }
+            }
+        }
+
+        let mut by_hour: Vec<HourPerformance> = hourly
+            .into_iter()
+            .map(|(hour, (trades, wins, total_pnl))| HourPerformance {
+                hour,
+                trades,
+                wins,
+                total_pnl: (total_pnl * 100.0).round() / 100.0,
+                win_rate: if trades > 0 { (wins as f64 / trades as f64 * 1000.0).round() / 10.0 } else { 0.0 },
+            })
+            .collect();
+        by_hour.sort_by_key(|h| h.hour);
+
+        let mut by_day: Vec<DayPerformance> = daily
+            .into_iter()
+            .map(|(day, (trades, wins, total_pnl, day_num))| DayPerformance {
+                day: day.clone(),
+                day_num,
+                trades,
+                wins,
+                total_pnl: (total_pnl * 100.0).round() / 100.0,
+                win_rate: if trades > 0 { (wins as f64 / trades as f64 * 1000.0).round() / 10.0 } else { 0.0 },
+            })
+            .collect();
+        by_day.sort_by_key(|d| d.day_num);
+
+        let best_hour = by_hour.iter().max_by(|a, b| a.total_pnl.partial_cmp(&b.total_pnl).unwrap()).map(|h| h.hour).unwrap_or(-1);
+        let worst_hour = by_hour.iter().min_by(|a, b| a.total_pnl.partial_cmp(&b.total_pnl).unwrap()).map(|h| h.hour).unwrap_or(-1);
+        let best_day = by_day.iter().max_by(|a, b| a.total_pnl.partial_cmp(&b.total_pnl).unwrap()).map(|d| d.day.clone()).unwrap_or_default();
+        let worst_day = by_day.iter().min_by(|a, b| a.total_pnl.partial_cmp(&b.total_pnl).unwrap()).map(|d| d.day.clone()).unwrap_or_default();
+
+        TimePerformance {
+            by_hour,
+            by_day,
+            best_hour,
+            worst_hour,
+            best_day,
+            worst_day,
+        }
+    }
+
+    pub fn hold_time_analysis(&self, deals: &[Deal]) -> HoldTimeAnalysis {
+        let mut hold_times: Vec<(f64, f64)> = Vec::new(); // (hold_minutes, profit)
+        let mut open_pos: HashMap<String, DateTime<chrono::Utc>> = HashMap::new();
+
+        for deal in deals {
+            let entry = deal.entry.to_lowercase();
+            if let Some(dt) = Self::parse_datetime(&deal.time) {
+                if entry.contains("in") && !entry.contains("out") {
+                    open_pos.insert(deal.order.clone(), dt);
+                } else if entry.contains("out") && deal.profit != 0.0 {
+                    if let Some(in_time) = open_pos.remove(&deal.order) {
+                        let hold_minutes = (dt - in_time).num_seconds() as f64 / 60.0;
+                        if hold_minutes > 0.0 {
+                            hold_times.push((hold_minutes, deal.profit));
+                        }
+                    }
+                }
+            }
+        }
+
+        if hold_times.is_empty() {
+            return HoldTimeAnalysis {
+                avg_hold_minutes: 0.0,
+                median_hold_minutes: 0.0,
+                buckets: vec![],
+                correlation_with_profit: 0.0,
+            };
+        }
+
+        let avg_hold = hold_times.iter().map(|(h, _)| *h).sum::<f64>() / hold_times.len() as f64;
+        let mut sorted_hold: Vec<f64> = hold_times.iter().map(|(h, _)| *h).collect();
+        sorted_hold.sort_by(|a, b| a.partial_cmp(b).unwrap());
+        let median_hold = sorted_hold[sorted_hold.len() / 2];
+
+        // Calculate correlation
+        let n = hold_times.len() as f64;
+        let sum_x = hold_times.iter().map(|(h, _)| *h).sum::<f64>();
+        let sum_y = hold_times.iter().map(|(_, p)| *p).sum::<f64>();
+        let sum_xy = hold_times.iter().map(|(h, p)| h * p).sum::<f64>();
+        let sum_x2 = hold_times.iter().map(|(h, _)| h * h).sum::<f64>();
+        let sum_y2 = hold_times.iter().map(|(_, p)| p * p).sum::<f64>();
+
+        let correlation = if n > 1.0 {
+            let numerator = n * sum_xy - sum_x * sum_y;
+            let denominator = ((n * sum_x2 - sum_x * sum_x) * (n * sum_y2 - sum_y * sum_y)).sqrt();
+            if denominator > 0.0 { numerator / denominator } else { 0.0 }
+        } else {
+            0.0
+        };
+
+        // Create buckets
+        let bucket_defs = [
+            (0.0, 15.0, "< 15 min"),
+            (15.0, 60.0, "15-60 min"),
+            (60.0, 240.0, "1-4 hours"),
+            (240.0, 1440.0, "4-24 hours"),
+            (1440.0, 10080.0, "1-7 days"),
+            (10080.0, 999999.0, "> 7 days"),
+        ];
+
+        let buckets: Vec<HoldTimeBucket> = bucket_defs
+            .iter()
+            .map(|(min, max, range)| {
+                let bucket_deals: Vec<(f64, f64)> = hold_times
+                    .iter()
+                    .filter(|(h, _)| *h >= *min && *h < *max)
+                    .cloned()
+                    .collect();
+                let count = bucket_deals.len() as i32;
+                let total_pnl: f64 = bucket_deals.iter().map(|(_, p)| *p).sum();
+                let avg_profit = if count > 0 { total_pnl / count as f64 } else { 0.0 };
+
+                HoldTimeBucket {
+                    range: range.to_string(),
+                    min_minutes: *min,
+                    max_minutes: *max,
+                    count,
+                    avg_profit: (avg_profit * 100.0).round() / 100.0,
+                    total_pnl: (total_pnl * 100.0).round() / 100.0,
+                }
+            })
+            .collect();
+
+        HoldTimeAnalysis {
+            avg_hold_minutes: (avg_hold * 10.0).round() / 10.0,
+            median_hold_minutes: (median_hold * 10.0).round() / 10.0,
+            buckets,
+            correlation_with_profit: (correlation * 1000.0).round() / 1000.0,
+        }
+    }
+
+    pub fn layer_performance(&self, deals: &[Deal]) -> Vec<LayerPerformance> {
+        let mut layer_stats: HashMap<i32, (i32, i32, f64, f64)> = HashMap::new();
+
+        for deal in deals {
+            let entry = deal.entry.to_lowercase();
+            if entry.contains("out") && deal.profit != 0.0 {
+                let layer = self.extract_layer(&deal.comment);
+                let stats = layer_stats.entry(layer).or_insert((0, 0, 0.0, 0.0));
+                stats.0 += 1;
+                stats.2 += deal.profit;
+                stats.3 += deal.volume;
+                if deal.profit > 0.0 {
+                    stats.1 += 1;
+                }
+            }
+        }
+
+        let mut result: Vec<LayerPerformance> = layer_stats
+            .into_iter()
+            .map(|(layer, (trades, wins, total_pnl, total_volume))| LayerPerformance {
+                layer,
+                trades,
+                wins,
+                total_pnl: (total_pnl * 100.0).round() / 100.0,
+                win_rate: if trades > 0 { (wins as f64 / trades as f64 * 1000.0).round() / 10.0 } else { 0.0 },
+                avg_volume: if trades > 0 { (total_volume / trades as f64 * 10000.0).round() / 10000.0 } else { 0.0 },
+                avg_profit: if trades > 0 { (total_pnl / trades as f64 * 100.0).round() / 100.0 } else { 0.0 },
+            })
+            .collect();
+
+        result.sort_by_key(|l| l.layer);
+        result
+    }
+
+    pub fn volume_analysis(&self, deals: &[Deal]) -> VolumeAnalysis {
+        let closed: Vec<&Deal> = deals
+            .iter()
+            .filter(|d| d.entry.to_lowercase().contains("out") && d.profit != 0.0)
+            .collect();
+
+        if closed.is_empty() {
+            return VolumeAnalysis {
+                correlation_with_profit: 0.0,
+                by_volume_bucket: vec![],
+            };
+        }
+
+        // Calculate correlation
+        let n = closed.len() as f64;
+        let sum_x: f64 = closed.iter().map(|d| d.volume).sum();
+        let sum_y: f64 = closed.iter().map(|d| d.profit).sum();
+        let sum_xy: f64 = closed.iter().map(|d| d.volume * d.profit).sum();
+        let sum_x2: f64 = closed.iter().map(|d| d.volume * d.volume).sum();
+        let sum_y2: f64 = closed.iter().map(|d| d.profit * d.profit).sum();
+
+        let correlation = if n > 1.0 {
+            let numerator = n * sum_xy - sum_x * sum_y;
+            let denominator = ((n * sum_x2 - sum_x * sum_x) * (n * sum_y2 - sum_y * sum_y)).sqrt();
+            if denominator > 0.0 { numerator / denominator } else { 0.0 }
+        } else {
+            0.0
+        };
+
+        // Create volume buckets
+        let bucket_defs = [
+            (0.0, 0.1, "0.0-0.1 lots"),
+            (0.1, 0.5, "0.1-0.5 lots"),
+            (0.5, 1.0, "0.5-1.0 lots"),
+            (1.0, 2.0, "1.0-2.0 lots"),
+            (2.0, 5.0, "2.0-5.0 lots"),
+            (5.0, 999.0, "5.0+ lots"),
+        ];
+
+        let by_volume_bucket: Vec<VolumeBucket> = bucket_defs
+            .iter()
+            .map(|(min, max, range)| {
+                let bucket_deals: Vec<&Deal> = closed
+                    .iter()
+                    .filter(|d| d.volume >= *min && d.volume < *max)
+                    .cloned()
+                    .collect();
+                let trades = bucket_deals.len() as i32;
+                let total_pnl: f64 = bucket_deals.iter().map(|d| d.profit).sum();
+                let avg_profit = if trades > 0 { total_pnl / trades as f64 } else { 0.0 };
+
+                VolumeBucket {
+                    volume_range: range.to_string(),
+                    min_volume: *min,
+                    max_volume: *max,
+                    trades,
+                    avg_profit: (avg_profit * 100.0).round() / 100.0,
+                    total_pnl: (total_pnl * 100.0).round() / 100.0,
+                }
+            })
+            .collect();
+
+        VolumeAnalysis {
+            correlation_with_profit: (correlation * 1000.0).round() / 1000.0,
+            by_volume_bucket,
+        }
+    }
+
+    pub fn cost_analysis(&self, deals: &[Deal]) -> CostAnalysis {
+        let total_commission: f64 = deals.iter().map(|d| d.commission.abs()).sum();
+        let total_swap: f64 = deals.iter().map(|d| d.swap.abs()).sum();
+        let gross_profit: f64 = deals.iter().map(|d| d.profit).filter(|p| *p > 0.0).sum();
+        let trade_count = deals.iter().filter(|d| d.entry.to_lowercase().contains("out")).count() as f64;
+
+        let commission_pct = if gross_profit > 0.0 { (total_commission / gross_profit * 10000.0).round() / 100.0 } else { 0.0 };
+        let swap_pct = if gross_profit > 0.0 { (total_swap / gross_profit * 10000.0).round() / 100.0 } else { 0.0 };
+
+        // Calculate what win rate would be without costs
+        let wins_before_costs = deals
+            .iter()
+            .filter(|d| {
+                let profit_before_costs = d.profit + d.commission.abs() + d.swap.abs();
+                d.entry.to_lowercase().contains("out") && profit_before_costs > 0.0
+            })
+            .count() as f64;
+        let total_closed = deals.iter().filter(|d| d.entry.to_lowercase().contains("out")).count() as f64;
+        let win_rate_before_costs = if total_closed > 0.0 { wins_before_costs / total_closed * 100.0 } else { 0.0 };
+
+        let current_wins = deals.iter().filter(|d| d.entry.to_lowercase().contains("out") && d.profit > 0.0).count() as f64;
+        let current_win_rate = if total_closed > 0.0 { current_wins / total_closed * 100.0 } else { 0.0 };
+
+        CostAnalysis {
+            total_commission: (total_commission * 100.0).round() / 100.0,
+            total_swap: (total_swap * 100.0).round() / 100.0,
+            commission_pct_of_profit: commission_pct,
+            swap_pct_of_profit: swap_pct,
+            avg_commission_per_trade: if trade_count > 0.0 { (total_commission / trade_count * 100.0).round() / 100.0 } else { 0.0 },
+            avg_swap_per_trade: if trade_count > 0.0 { (total_swap / trade_count * 100.0).round() / 100.0 } else { 0.0 },
+            net_profit_before_costs: (gross_profit * 100.0).round() / 100.0,
+            cost_impact_on_win_rate: (win_rate_before_costs - current_win_rate * 100.0).round() / 100.0,
+        }
+    }
+
+    pub fn efficiency_analysis(&self, deals: &[Deal], _metrics: &Metrics) -> EfficiencyAnalysis {
+        let closed: Vec<&Deal> = deals
+            .iter()
+            .filter(|d| d.entry.to_lowercase().contains("out") && d.profit != 0.0)
+            .collect();
+
+        if closed.is_empty() {
+            return EfficiencyAnalysis {
+                profit_per_hour: 0.0,
+                profit_per_day: 0.0,
+                profit_per_trade_hour: 0.0,
+                avg_trade_duration_hours: 0.0,
+                annualized_return_pct: 0.0,
+                trades_per_day: 0.0,
+            };
+        }
+
+        let total_profit: f64 = closed.iter().map(|d| d.profit).sum();
+        let total_trades = closed.len() as f64;
+
+        // Calculate total hold time
+        let mut total_hold_minutes = 0.0;
+        let mut open_pos: HashMap<String, DateTime<chrono::Utc>> = HashMap::new();
+
+        for deal in deals {
+            let entry = deal.entry.to_lowercase();
+            if let Some(dt) = Self::parse_datetime(&deal.time) {
+                if entry.contains("in") && !entry.contains("out") {
+                    open_pos.insert(deal.order.clone(), dt);
+                } else if entry.contains("out") && deal.profit != 0.0 {
+                    if let Some(in_time) = open_pos.remove(&deal.order) {
+                        let hold_minutes = (dt - in_time).num_seconds() as f64 / 60.0;
+                        if hold_minutes > 0.0 {
+                            total_hold_minutes += hold_minutes;
+                        }
+                    }
+                }
+            }
+        }
+
+        let total_hold_hours = total_hold_minutes / 60.0;
+        let avg_trade_duration = if total_trades > 0.0 { total_hold_minutes / total_trades / 60.0 } else { 0.0 };
+
+        // Get date range
+        let dates: Vec<DateTime<chrono::Utc>> = deals
+            .iter()
+            .filter_map(|d| Self::parse_datetime(&d.time))
+            .collect();
+
+        let total_days = if dates.len() >= 2 {
+            let min_date = dates.iter().min().unwrap();
+            let max_date = dates.iter().max().unwrap();
+            (*max_date - *min_date).num_days().max(1) as f64
+        } else {
+            1.0
+        };
+
+        // Use a default deposit of 10000 for annualized calculation
+        // In real scenarios, this should come from the report
+        let deposit = 10000.0;
+        let annualized = if total_days > 0.0 && deposit > 0.0 {
+            let daily_return = total_profit / deposit / total_days;
+            ((1.0 + daily_return).powf(365.0) - 1.0) * 100.0
+        } else {
+            0.0
+        };
+
+        EfficiencyAnalysis {
+            profit_per_hour: if total_hold_hours > 0.0 { (total_profit / total_hold_hours * 100.0).round() / 100.0 } else { 0.0 },
+            profit_per_day: (total_profit / total_days * 100.0).round() / 100.0,
+            profit_per_trade_hour: if total_hold_hours > 0.0 { (total_profit / total_hold_hours / total_trades * 100.0).round() / 100.0 } else { 0.0 },
+            avg_trade_duration_hours: (avg_trade_duration * 10.0).round() / 10.0,
+            annualized_return_pct: (annualized * 10.0).round() / 10.0,
+            trades_per_day: (total_trades / total_days * 10.0).round() / 10.0,
+        }
+    }
+}
@@ -1,7 +1,8 @@
 use anyhow::{anyhow, Result};
 use std::fs;
 use std::path::{Path, PathBuf};
-use std::process::Command;
+use std::time::Duration;
+use tokio::time::timeout as tokio_timeout;

 use crate::models::Config;

@@ -28,7 +29,11 @@ impl MqlCompiler {
        Self { config }
    }

-    pub fn compile(&self, source_path: &str) -> Result<CompileResult> {
+    pub async fn compile(&self, source_path: &str) -> Result<CompileResult> {
+        self.compile_with_timeout(source_path, Duration::from_secs(120)).await
+    }
+
+    pub async fn compile_with_timeout(&self, source_path: &str, timeout: Duration) -> Result<CompileResult> {
        let source_path = Path::new(source_path);
        if !source_path.exists() {
            return Err(anyhow!("Source file not found: {}", source_path.display()));
@@ -61,7 +66,7 @@ impl MqlCompiler {
        let staged_mq5 = &sync.dest_mq5;
        tracing::info!("Staged {} file(s) to: {}", sync.files_copied, staged_mq5.display());

-        self.run_metaeditor(wine_exe, &wine_prefix, &metaeditor, staged_mq5)?;
+        self.run_metaeditor_with_timeout(wine_exe, &wine_prefix, &metaeditor, staged_mq5, timeout).await?;

        // /log flag (no path) writes log adjacent to source: {ea_name}.log
        let log_path = staged_mq5.with_extension("log");
@@ -197,15 +202,16 @@ impl MqlCompiler {
        Ok(SyncStats { dest_mq5, files_copied })
    }

-    /// Run MetaEditor to compile `source_mq5`.
+    /// Run MetaEditor to compile `source_mq5` with timeout.
    /// Uses Unix host path for /compile: and bare /log flag (writes log adjacent to source).
    /// Shell script intermediary required on macOS to preserve DYLD_* vars past SIP.
-    fn run_metaeditor(
+    async fn run_metaeditor_with_timeout(
        &self,
        wine_exe: &str,
        wine_prefix: &Path,
        metaeditor: &Path,
        source_mq5: &Path,
+        timeout: Duration,
    ) -> Result<()> {
        let mt5_dir = metaeditor.parent().unwrap_or(metaeditor);

@@ -242,16 +248,24 @@ impl MqlCompiler {
                use std::os::unix::fs::PermissionsExt;
                fs::set_permissions(&script_path, fs::Permissions::from_mode(0o755))?;
            }
-            Command::new("/bin/sh").arg(&script_path).output()?;
+            let compile_future = tokio::process::Command::new("/bin/sh")
+                .arg(&script_path)
+                .output();
+            let result = tokio_timeout(timeout, compile_future).await
+                .map_err(|_| anyhow!("Compilation timed out after {} seconds", timeout.as_secs()))?;
+            result?;
        } else {
-            Command::new(wine_exe)
+            let compile_future = tokio::process::Command::new(wine_exe)
                .arg(metaeditor)
                .arg(format!("/compile:{}", source_mq5.display()))
                .arg("/log")
                .env("WINEPREFIX", wine_prefix)
                .env("WINEDEBUG", "-all")
                .current_dir(mt5_dir)
-                .output()?;
+                .output();
+            let result = tokio_timeout(timeout, compile_future).await
+                .map_err(|_| anyhow!("Compilation timed out after {} seconds", timeout.as_secs()))?;
+            result?;
        }
        Ok(())
    }
@@ -40,8 +40,11 @@ pub struct McpRequest {
 #[derive(Debug, serde::Deserialize, serde::Serialize)]
 pub struct McpResponse {
    jsonrpc: String,
+    #[serde(skip_serializing_if = "Option::is_none")]
    id: Option<Value>,
+    #[serde(skip_serializing_if = "Option::is_none")]
    result: Option<Value>,
+    #[serde(skip_serializing_if = "Option::is_none")]
    error: Option<McpError>,
 }

@@ -49,6 +52,7 @@ pub struct McpResponse {
 pub struct McpError {
    code: i32,
    message: String,
+    #[serde(skip_serializing_if = "Option::is_none")]
    data: Option<Value>,
 }

@@ -78,7 +82,9 @@ pub struct ToolCapabilities {

 #[tokio::main]
 async fn main() -> Result<()> {
-    tracing_subscriber::fmt::init();
+    tracing_subscriber::fmt()
+        .with_writer(std::io::stderr)
+        .init();
    
    let cli = Cli::parse();
    
@@ -111,6 +117,11 @@ async fn run_stdio_server() -> Result<()> {
                let server_clone = server.clone();
                match serde_json::from_str::<McpRequest>(line) {
                    Ok(request) => {
+                        // Notifications have no id — don't send a response
+                        if request.id.is_none() {
+                            server_clone.handle_notification(request).await;
+                            continue;
+                        }
                        let response = server_clone.handle_request(request).await;
                        let response_json = serde_json::to_string(&response)?;
                        println!("{}", response_json);
@@ -180,6 +191,10 @@ async fn handle_connection(socket: tokio::net::TcpStream) -> Result<()> {
                
                match serde_json::from_str::<McpRequest>(line) {
                    Ok(request) => {
+                        if request.id.is_none() {
+                            server.handle_notification(request).await;
+                            continue;
+                        }
                        let response = server.handle_request(request).await;
                        let response_json = serde_json::to_string(&response)? + "\n";
                        writer.write_all(response_json.as_bytes()).await?;
@@ -99,6 +99,18 @@ impl McpServer {
        }
    }

+    /// Handle a notification (no id — no response sent)
+    pub async fn handle_notification(&self, request: McpRequest) {
+        match request.method.as_str() {
+            "notifications/initialized" => {
+                // Client confirms initialization is complete — no action needed
+            }
+            _ => {
+                tracing::debug!("Unhandled notification: {}", request.method);
+            }
+        }
+    }
+
    pub async fn handle_request(&self, request: McpRequest) -> McpResponse {
        match request.method.as_str() {
            "initialize" => {
@@ -123,9 +135,8 @@ impl McpServer {
                // Return immediately with fast status
                let server_info = json!({
                    "name": "MT5-Quant",
-                    "version": "1.27.0",
+                    "version": env!("CARGO_PKG_VERSION"),
                    "setup": {
-                        "verified": null,  // null = checking
                        "hint": "Auto-verification running... Use verify_setup tool for detailed status",
                    }
                });
@@ -164,7 +175,7 @@ impl McpServer {
                McpResponse {
                    jsonrpc: "2.0".to_string(),
                    id: request.id,
-                    result: Some(crate::tools::get_tools_list()),
+                    result: Some(json!({"tools": crate::tools::get_tools_list()})),
                    error: None,
                }
            }
@@ -1,6 +1,9 @@
 use serde::{Deserialize, Serialize};
 use std::path::PathBuf;

+// Re-export chrono for BacktestJob
+use chrono;
+
 #[allow(dead_code)]
 #[derive(Debug, Clone, Serialize, Deserialize)]
 pub struct Report {
@@ -31,6 +34,8 @@ pub struct PipelineMetadata {
    pub report_dir: String,
    pub duration_seconds: i64,
    pub files: FilePaths,
+    #[serde(default)]
+    pub no_trades: bool,
 }

 #[derive(Debug, Clone, Serialize, Deserialize)]
@@ -48,6 +53,9 @@ pub struct BacktestStatus {
    pub elapsed_seconds: i64,
    pub is_complete: bool,
    pub message: String,
+    pub report_dir: Option<String>,
+    pub mt5_running: Option<bool>,
+    pub report_found: Option<bool>,
 }

 #[allow(dead_code)]
@@ -60,4 +68,43 @@ pub enum PipelineStage {
    Extract,
    Analyze,
    Done,
+    Failed,
+}
+
+/// Track a running backtest job for fire-and-poll pattern
+#[derive(Debug, Clone, Serialize, Deserialize)]
+pub struct BacktestJob {
+    pub report_id: String,
+    pub report_dir: String,
+    pub expert: String,
+    pub symbol: String,
+    pub timeframe: String,
+    pub launched_at: String,
+    pub mt5_pid: Option<u32>,
+    pub expected_report_path: String,
+    pub timeout_seconds: u64,
+}
+
+impl BacktestJob {
+    pub fn new(
+        report_id: String,
+        report_dir: String,
+        expert: String,
+        symbol: String,
+        timeframe: String,
+        expected_report_path: String,
+        timeout_seconds: u64,
+    ) -> Self {
+        Self {
+            report_id,
+            report_dir,
+            expert,
+            symbol,
+            timeframe,
+            launched_at: chrono::Utc::now().to_rfc3339(),
+            mt5_pid: None,
+            expected_report_path,
+            timeout_seconds,
+        }
+    }
 }
@@ -8,7 +8,7 @@ use tokio::time::{sleep, Duration};
 use crate::analytics::{DealAnalyzer, ReportExtractor};
 use crate::compile::MqlCompiler;
 use crate::models::config::Config;
-use crate::models::report::{PipelineMetadata, FilePaths};
+use crate::models::report::{PipelineMetadata, FilePaths, BacktestJob};
 use crate::storage::{ReportDb, ReportEntry};

 pub struct BacktestPipeline {
@@ -73,7 +73,7 @@ impl BacktestPipeline {

        if !params.skip_compile {
            self.log_progress(&progress_log, "COMPILE").await;
-            self.compile_ea(&params.expert).await?;
+            self.compile_ea(&params.expert, params.timeout).await?;
        }

        if !params.skip_clean {
@@ -90,6 +90,13 @@ impl BacktestPipeline {
            &report_dir.to_string_lossy(),
        )?;

+        // Handle case where EA didn't trade - no deals generated
+        if extraction.deals.is_empty() {
+            tracing::warn!("Backtest completed but no deals were generated - EA did not trade during this period");
+            let warning_path = report_dir.join("NO_TRADES_WARNING.txt");
+            let _ = fs::write(&warning_path, "Warning: No deals were generated during this backtest.\nThe EA did not execute any trades during the specified date range.\n");
+        }
+
        // Move equity chart images to OS temp dir, then delete the HTML report.
        let charts_dir = self.relocate_charts(&report_path, &report_id).await;
        let _ = fs::remove_file(&report_path);
@@ -108,7 +115,7 @@ impl BacktestPipeline {
        self.log_progress(&progress_log, "DONE").await;

        let duration = (chrono::Utc::now() - start_time).num_seconds();
-        self.save_metadata(&params, &report_dir, duration).await?;
+        self.save_metadata(&params, &report_dir, duration, extraction.deals.is_empty()).await?;

        // Register in the SQLite report registry.
        self.register_in_db(
@@ -122,14 +129,105 @@ impl BacktestPipeline {
        )
        .await;

+        let message = if extraction.deals.is_empty() {
+            "Backtest completed successfully, but EA did not execute any trades during this period".to_string()
+        } else {
+            "Backtest completed successfully".to_string()
+        };
+
        Ok(PipelineResult {
            success: true,
            report_dir,
            duration_seconds: duration,
-            message: "Backtest completed successfully".to_string(),
+            message,
        })
    }

+    /// Launch backtest in fire-and-forget mode: compile, clean, launch MT5, return immediately.
+    /// Returns a BacktestJob that can be used with get_backtest_status to poll for completion.
+    pub async fn launch_backtest(&self, params: BacktestParams) -> Result<BacktestJob> {
+        let _start_time = chrono::Utc::now();
+        let report_id = self.generate_report_id(&params);
+        let report_dir = self.config.reports_dir().join(&report_id);
+
+        fs::create_dir_all(&report_dir)?;
+
+        let progress_log = report_dir.join("progress.log");
+        self.log_progress(&progress_log, "START").await;
+
+        if !params.skip_compile {
+            self.log_progress(&progress_log, "COMPILE").await;
+            self.compile_ea(&params.expert, params.timeout).await?;
+        }
+
+        if !params.skip_clean {
+            self.log_progress(&progress_log, "CLEAN").await;
+            self.clean_cache(&params.expert).await?;
+        }
+
+        self.log_progress(&progress_log, "BACKTEST").await;
+        
+        // Get MT5 paths
+        let mt5_dir = self.config.mt5_dir()
+            .ok_or_else(|| anyhow!("MT5 directory not configured"))?;
+        let wine_exe = self.config.wine_executable.as_ref()
+            .ok_or_else(|| anyhow!("wine_executable not configured"))?;
+        let wine_prefix = mt5_dir
+            .parent()
+            .and_then(|p| p.parent())
+            .and_then(|p| p.parent())
+            .map(|p| p.to_path_buf())
+            .ok_or_else(|| anyhow!("Could not determine Wine prefix from terminal_dir"))?;
+        let reports_dir = mt5_dir.join("reports");
+        fs::create_dir_all(&reports_dir)?;
+
+        // Write config files
+        let ini_content = self.build_backtest_ini(&params, &report_id)?;
+        let config_host = wine_prefix.join("drive_c").join("backtest_config.ini");
+        fs::write(&config_host, ini_content.as_bytes())?;
+        self.update_terminal_ini(&params, &report_id)?;
+
+        // Kill any running MT5
+        self.kill_mt5().await?;
+
+        // Launch MT5 (fire and forget)
+        let mut cmd = self.build_wine_launch(wine_exe, &wine_prefix)?;
+        let child = cmd.stdin(std::process::Stdio::null())
+            .stdout(std::process::Stdio::null())
+            .stderr(std::process::Stdio::null())
+            .spawn()?;
+
+        let pid = child.id();
+        tracing::info!("MT5 launched with PID {:?} for backtest {}", pid, report_id);
+
+        // Create and save the job tracking file
+        let expected_report = reports_dir.join(format!("{}.htm", report_id));
+        let job = BacktestJob::new(
+            report_id.clone(),
+            report_dir.to_string_lossy().to_string(),
+            params.expert.clone(),
+            params.symbol.clone(),
+            params.timeframe.clone(),
+            expected_report.to_string_lossy().to_string(),
+            params.timeout,
+        );
+
+        // Save job info for polling
+        let job_path = report_dir.join("job.json");
+        fs::write(&job_path, serde_json::to_string_pretty(&job)?)?;
+
+        // Save initial metadata
+        self.save_metadata(&params, &report_dir, 0, false).await?;
+
+        // Register in DB as "running"
+        let db = ReportDb::new(&Config::db_path());
+        if let Err(e) = db.init() {
+            tracing::warn!("Failed to init report DB: {}", e);
+        }
+
+        Ok(job)
+    }
+
    /// Move equity chart images (*.png, *.gif) from MT5's reports dir to OS temp,
    /// returning the temp path if any images were found.
    async fn relocate_charts(&self, html_path: &Path, report_id: &str) -> Option<PathBuf> {
@@ -232,7 +330,7 @@ impl BacktestPipeline {
        }
    }

-    async fn compile_ea(&self, expert: &str) -> Result<()> {
+    async fn compile_ea(&self, expert: &str, timeout_secs: u64) -> Result<()> {
        let mut search_paths = vec![
            PathBuf::from(&self.config.get("project_dir")).join("src/experts").join(format!("{}.mq5", expert)),
            PathBuf::from(&self.config.get("project_dir")).join("src").join(format!("{}.mq5", expert)),
@@ -252,7 +350,8 @@ impl BacktestPipeline {
            .find(|p| p.exists())
            .ok_or_else(|| anyhow!("Cannot find {}.mq5 — searched project_dir and MT5 Experts dir", expert))?;

-        let result = self.compiler.compile(&source_path.to_string_lossy())?;
+        let timeout = std::time::Duration::from_secs(timeout_secs.min(300)); // Max 5 min for compile
+        let result = self.compiler.compile_with_timeout(&source_path.to_string_lossy(), timeout).await?;
        
        if !result.success {
            return Err(anyhow!(
@@ -761,7 +860,7 @@ impl BacktestPipeline {
        let _ = fs::write(log_path, line);
    }

-    async fn save_metadata(&self, params: &BacktestParams, report_dir: &Path, duration: i64) -> Result<()> {
+    async fn save_metadata(&self, params: &BacktestParams, report_dir: &Path, duration: i64, no_trades: bool) -> Result<()> {
        let metadata = PipelineMetadata {
            expert: params.expert.clone(),
            symbol: params.symbol.clone(),
@@ -781,6 +880,7 @@ impl BacktestPipeline {
                deals_csv: report_dir.join("deals.csv").to_string_lossy().to_string(),
                deals_json: report_dir.join("deals.json").to_string_lossy().to_string(),
            },
+            no_trades,
        };

        let json = serde_json::to_string_pretty(&metadata)?;
@@ -374,4 +374,543 @@ impl ReportDb {

        Ok(entry)
    }
+
+    /// Get a specific report by ID
+    pub fn get_by_id(&self, id: &str) -> Result<Option<ReportEntry>> {
+        let conn = self.connect()?;
+        let mut stmt = conn.prepare(
+            "SELECT id, expert, symbol, timeframe, model, from_date, to_date, \
+            created_at, set_file_original, set_snapshot_path, report_dir, charts_dir, \
+            net_profit, profit_factor, max_dd_pct, sharpe_ratio, total_trades, \
+            win_rate_pct, recovery_factor, deposit, currency, leverage, \
+            duration_seconds, tags, notes, verdict \
+            FROM reports WHERE id = ?"
+        )?;
+
+        let entry = stmt
+            .query_map([id], |row| {
+                let tags_json: String = row.get(23)?;
+                let tags: Vec<String> = serde_json::from_str(&tags_json).unwrap_or_default();
+                Ok(ReportEntry {
+                    id: row.get(0)?,
+                    expert: row.get(1)?,
+                    symbol: row.get(2)?,
+                    timeframe: row.get(3)?,
+                    model: row.get(4)?,
+                    from_date: row.get(5)?,
+                    to_date: row.get(6)?,
+                    created_at: row.get(7)?,
+                    set_file_original: row.get(8)?,
+                    set_snapshot_path: row.get(9)?,
+                    report_dir: row.get(10)?,
+                    charts_dir: row.get(11)?,
+                    net_profit: row.get(12)?,
+                    profit_factor: row.get(13)?,
+                    max_dd_pct: row.get(14)?,
+                    sharpe_ratio: row.get(15)?,
+                    total_trades: row.get(16)?,
+                    win_rate_pct: row.get(17)?,
+                    recovery_factor: row.get(18)?,
+                    deposit: row.get(19)?,
+                    currency: row.get(20)?,
+                    leverage: row.get(21)?,
+                    duration_seconds: row.get(22)?,
+                    tags,
+                    notes: row.get(24)?,
+                    verdict: row.get(25)?,
+                })
+            })?
+            .filter_map(|r| r.ok())
+            .next();
+
+        Ok(entry)
+    }
+
+    /// Search reports by tags (at least one tag must match)
+    pub fn search_by_tags(&self, tags: &[String], limit: usize) -> Result<Vec<ReportEntry>> {
+        let conn = self.connect()?;
+        let mut sql = "SELECT id, expert, symbol, timeframe, model, from_date, to_date, \
+            created_at, set_file_original, set_snapshot_path, report_dir, charts_dir, \
+            net_profit, profit_factor, max_dd_pct, sharpe_ratio, total_trades, \
+            win_rate_pct, recovery_factor, deposit, currency, leverage, \
+            duration_seconds, tags, notes, verdict \
+            FROM reports WHERE 1=1"
+            .to_string();
+
+        // Build OR conditions for tags - use JSON1 extension for tag matching
+        if !tags.is_empty() {
+            let tag_conditions: Vec<String> = tags.iter()
+                .map(|tag| format!("tags LIKE '%{}%'", tag.replace("'", "''")))
+                .collect();
+            sql.push_str(&format!(" AND ({})", tag_conditions.join(" OR ")));
+        }
+        sql.push_str(&format!(" ORDER BY created_at DESC LIMIT {}", limit));
+
+        let mut stmt = conn.prepare(&sql)?;
+        let entries: Vec<ReportEntry> = stmt
+            .query_map([], |row| {
+                let tags_str: String = row.get(23).unwrap_or_else(|_| "[]".to_string());
+                let tags: Vec<String> = serde_json::from_str(&tags_str).unwrap_or_default();
+                Ok(ReportEntry {
+                    id: row.get(0)?,
+                    expert: row.get(1)?,
+                    symbol: row.get(2)?,
+                    timeframe: row.get(3)?,
+                    model: row.get(4)?,
+                    from_date: row.get(5)?,
+                    to_date: row.get(6)?,
+                    created_at: row.get(7)?,
+                    set_file_original: row.get(8)?,
+                    set_snapshot_path: row.get(9)?,
+                    report_dir: row.get(10)?,
+                    charts_dir: row.get(11)?,
+                    net_profit: row.get(12)?,
+                    profit_factor: row.get(13)?,
+                    max_dd_pct: row.get(14)?,
+                    sharpe_ratio: row.get(15)?,
+                    total_trades: row.get(16)?,
+                    win_rate_pct: row.get(17)?,
+                    recovery_factor: row.get(18)?,
+                    deposit: row.get(19)?,
+                    currency: row.get(20)?,
+                    leverage: row.get(21)?,
+                    duration_seconds: row.get(22)?,
+                    tags,
+                    notes: row.get(24)?,
+                    verdict: row.get(25)?,
+                })
+            })?
+            .filter_map(|r| r.ok())
+            .collect();
+
+        Ok(entries)
+    }
+
+    /// Search reports by notes text (case-insensitive LIKE)
+    pub fn search_by_notes(&self, query: &str, limit: usize) -> Result<Vec<ReportEntry>> {
+        let conn = self.connect()?;
+        let pattern = format!("%{}%", query.replace("'", "''"));
+        let mut stmt = conn.prepare(
+            &format!("SELECT id, expert, symbol, timeframe, model, from_date, to_date, \
+            created_at, set_file_original, set_snapshot_path, report_dir, charts_dir, \
+            net_profit, profit_factor, max_dd_pct, sharpe_ratio, total_trades, \
+            win_rate_pct, recovery_factor, deposit, currency, leverage, \
+            duration_seconds, tags, notes, verdict \
+            FROM reports WHERE notes LIKE '{}' ORDER BY created_at DESC LIMIT {}",
+            pattern, limit)
+        )?;
+
+        let entries: Vec<ReportEntry> = stmt
+            .query_map([], |row| {
+                let tags_str: String = row.get(23).unwrap_or_else(|_| "[]".to_string());
+                let tags: Vec<String> = serde_json::from_str(&tags_str).unwrap_or_default();
+                Ok(ReportEntry {
+                    id: row.get(0)?,
+                    expert: row.get(1)?,
+                    symbol: row.get(2)?,
+                    timeframe: row.get(3)?,
+                    model: row.get(4)?,
+                    from_date: row.get(5)?,
+                    to_date: row.get(6)?,
+                    created_at: row.get(7)?,
+                    set_file_original: row.get(8)?,
+                    set_snapshot_path: row.get(9)?,
+                    report_dir: row.get(10)?,
+                    charts_dir: row.get(11)?,
+                    net_profit: row.get(12)?,
+                    profit_factor: row.get(13)?,
+                    max_dd_pct: row.get(14)?,
+                    sharpe_ratio: row.get(15)?,
+                    total_trades: row.get(16)?,
+                    win_rate_pct: row.get(17)?,
+                    recovery_factor: row.get(18)?,
+                    deposit: row.get(19)?,
+                    currency: row.get(20)?,
+                    leverage: row.get(21)?,
+                    duration_seconds: row.get(22)?,
+                    tags,
+                    notes: row.get(24)?,
+                    verdict: row.get(25)?,
+                })
+            })?
+            .filter_map(|r| r.ok())
+            .collect();
+
+        Ok(entries)
+    }
+
+    /// Find reports by set file (original or snapshot)
+    pub fn search_by_set_file(&self, set_file: &str, limit: usize) -> Result<Vec<ReportEntry>> {
+        let conn = self.connect()?;
+        let pattern = format!("%{}%", set_file.replace("'", "''"));
+        let mut stmt = conn.prepare(
+            &format!("SELECT id, expert, symbol, timeframe, model, from_date, to_date, \
+            created_at, set_file_original, set_snapshot_path, report_dir, charts_dir, \
+            net_profit, profit_factor, max_dd_pct, sharpe_ratio, total_trades, \
+            win_rate_pct, recovery_factor, deposit, currency, leverage, \
+            duration_seconds, tags, notes, verdict \
+            FROM reports WHERE (set_file_original LIKE '{}' OR set_snapshot_path LIKE '{}') \
+            ORDER BY created_at DESC LIMIT {}",
+            pattern, pattern, limit)
+        )?;
+
+        let entries: Vec<ReportEntry> = stmt
+            .query_map([], |row| {
+                let tags_str: String = row.get(23).unwrap_or_else(|_| "[]".to_string());
+                let tags: Vec<String> = serde_json::from_str(&tags_str).unwrap_or_default();
+                Ok(ReportEntry {
+                    id: row.get(0)?,
+                    expert: row.get(1)?,
+                    symbol: row.get(2)?,
+                    timeframe: row.get(3)?,
+                    model: row.get(4)?,
+                    from_date: row.get(5)?,
+                    to_date: row.get(6)?,
+                    created_at: row.get(7)?,
+                    set_file_original: row.get(8)?,
+                    set_snapshot_path: row.get(9)?,
+                    report_dir: row.get(10)?,
+                    charts_dir: row.get(11)?,
+                    net_profit: row.get(12)?,
+                    profit_factor: row.get(13)?,
+                    max_dd_pct: row.get(14)?,
+                    sharpe_ratio: row.get(15)?,
+                    total_trades: row.get(16)?,
+                    win_rate_pct: row.get(17)?,
+                    recovery_factor: row.get(18)?,
+                    deposit: row.get(19)?,
+                    currency: row.get(20)?,
+                    leverage: row.get(21)?,
+                    duration_seconds: row.get(22)?,
+                    tags,
+                    notes: row.get(24)?,
+                    verdict: row.get(25)?,
+                })
+            })?
+            .filter_map(|r| r.ok())
+            .collect();
+
+        Ok(entries)
+    }
+
+    /// Search by backtest date range (from_date and to_date fields)
+    pub fn search_by_date_range(
+        &self,
+        from_start: Option<&str>,
+        from_end: Option<&str>,
+        to_start: Option<&str>,
+        to_end: Option<&str>,
+        limit: usize,
+    ) -> Result<Vec<ReportEntry>> {
+        let conn = self.connect()?;
+        let mut sql = "SELECT id, expert, symbol, timeframe, model, from_date, to_date, \
+            created_at, set_file_original, set_snapshot_path, report_dir, charts_dir, \
+            net_profit, profit_factor, max_dd_pct, sharpe_ratio, total_trades, \
+            win_rate_pct, recovery_factor, deposit, currency, leverage, \
+            duration_seconds, tags, notes, verdict \
+            FROM reports WHERE 1=1"
+            .to_string();
+
+        let mut params: Vec<String> = Vec::new();
+
+        if let Some(start) = from_start {
+            sql.push_str(" AND from_date >= ?");
+            params.push(start.to_string());
+        }
+        if let Some(end) = from_end {
+            sql.push_str(" AND from_date <= ?");
+            params.push(end.to_string());
+        }
+        if let Some(start) = to_start {
+            sql.push_str(" AND to_date >= ?");
+            params.push(start.to_string());
+        }
+        if let Some(end) = to_end {
+            sql.push_str(" AND to_date <= ?");
+            params.push(end.to_string());
+        }
+
+        sql.push_str(&format!(" ORDER BY created_at DESC LIMIT {}", limit));
+
+        let mut stmt = conn.prepare(&sql)?;
+        let entries: Vec<ReportEntry> = stmt
+            .query_map(rusqlite::params_from_iter(params.iter()), |row| {
+                let tags_str: String = row.get(23).unwrap_or_else(|_| "[]".to_string());
+                let tags: Vec<String> = serde_json::from_str(&tags_str).unwrap_or_default();
+                Ok(ReportEntry {
+                    id: row.get(0)?,
+                    expert: row.get(1)?,
+                    symbol: row.get(2)?,
+                    timeframe: row.get(3)?,
+                    model: row.get(4)?,
+                    from_date: row.get(5)?,
+                    to_date: row.get(6)?,
+                    created_at: row.get(7)?,
+                    set_file_original: row.get(8)?,
+                    set_snapshot_path: row.get(9)?,
+                    report_dir: row.get(10)?,
+                    charts_dir: row.get(11)?,
+                    net_profit: row.get(12)?,
+                    profit_factor: row.get(13)?,
+                    max_dd_pct: row.get(14)?,
+                    sharpe_ratio: row.get(15)?,
+                    total_trades: row.get(16)?,
+                    win_rate_pct: row.get(17)?,
+                    recovery_factor: row.get(18)?,
+                    deposit: row.get(19)?,
+                    currency: row.get(20)?,
+                    leverage: row.get(21)?,
+                    duration_seconds: row.get(22)?,
+                    tags,
+                    notes: row.get(24)?,
+                    verdict: row.get(25)?,
+                })
+            })?
+            .filter_map(|r| r.ok())
+            .collect();
+
+        Ok(entries)
+    }
+
+    /// Get comparable reports (same expert/symbol/timeframe for comparison)
+    pub fn get_comparable(
+        &self,
+        expert: &str,
+        symbol: &str,
+        timeframe: &str,
+        exclude_id: Option<&str>,
+        limit: usize,
+    ) -> Result<Vec<ReportEntry>> {
+        let conn = self.connect()?;
+        let mut sql = "SELECT id, expert, symbol, timeframe, model, from_date, to_date, \
+            created_at, set_file_original, set_snapshot_path, report_dir, charts_dir, \
+            net_profit, profit_factor, max_dd_pct, sharpe_ratio, total_trades, \
+            win_rate_pct, recovery_factor, deposit, currency, leverage, \
+            duration_seconds, tags, notes, verdict \
+            FROM reports WHERE expert = ? AND symbol = ? AND timeframe = ?"
+            .to_string();
+
+        let mut params: Vec<String> = vec![
+            expert.to_string(),
+            symbol.to_string(),
+            timeframe.to_string(),
+        ];
+
+        if let Some(id) = exclude_id {
+            sql.push_str(" AND id != ?");
+            params.push(id.to_string());
+        }
+
+        sql.push_str(&format!(" ORDER BY created_at DESC LIMIT {}", limit));
+
+        let mut stmt = conn.prepare(&sql)?;
+        let entries: Vec<ReportEntry> = stmt
+            .query_map(rusqlite::params_from_iter(params.iter()), |row| {
+                let tags_str: String = row.get(23).unwrap_or_else(|_| "[]".to_string());
+                let tags: Vec<String> = serde_json::from_str(&tags_str).unwrap_or_default();
+                Ok(ReportEntry {
+                    id: row.get(0)?,
+                    expert: row.get(1)?,
+                    symbol: row.get(2)?,
+                    timeframe: row.get(3)?,
+                    model: row.get(4)?,
+                    from_date: row.get(5)?,
+                    to_date: row.get(6)?,
+                    created_at: row.get(7)?,
+                    set_file_original: row.get(8)?,
+                    set_snapshot_path: row.get(9)?,
+                    report_dir: row.get(10)?,
+                    charts_dir: row.get(11)?,
+                    net_profit: row.get(12)?,
+                    profit_factor: row.get(13)?,
+                    max_dd_pct: row.get(14)?,
+                    sharpe_ratio: row.get(15)?,
+                    total_trades: row.get(16)?,
+                    win_rate_pct: row.get(17)?,
+                    recovery_factor: row.get(18)?,
+                    deposit: row.get(19)?,
+                    currency: row.get(20)?,
+                    leverage: row.get(21)?,
+                    duration_seconds: row.get(22)?,
+                    tags,
+                    notes: row.get(24)?,
+                    verdict: row.get(25)?,
+                })
+            })?
+            .filter_map(|r| r.ok())
+            .collect();
+
+        Ok(entries)
+    }
+
+    /// Get reports sorted by a specific metric (for best/worst queries)
+    pub fn get_sorted_by(
+        &self,
+        sort_column: &str,
+        ascending: bool,
+        limit: usize,
+        filters: &ReportFilters,
+    ) -> Result<Vec<ReportEntry>> {
+        let conn = self.connect()?;
+
+        // Validate sort column to prevent SQL injection
+        let valid_columns = ["net_profit", "profit_factor", "max_dd_pct", "win_rate_pct",
+                            "sharpe_ratio", "recovery_factor", "total_trades", "created_at"];
+        if !valid_columns.contains(&sort_column) {
+            return Err(anyhow::anyhow!("Invalid sort column: {}", sort_column));
+        }
+
+        let mut sql = "SELECT id, expert, symbol, timeframe, model, from_date, to_date, \
+            created_at, set_file_original, set_snapshot_path, report_dir, charts_dir, \
+            net_profit, profit_factor, max_dd_pct, sharpe_ratio, total_trades, \
+            win_rate_pct, recovery_factor, deposit, currency, leverage, \
+            duration_seconds, tags, notes, verdict \
+            FROM reports WHERE 1=1"
+            .to_string();
+
+        let mut filter_params: Vec<String> = Vec::new();
+
+        if let Some(ea) = &filters.expert {
+            sql.push_str(" AND expert LIKE ?");
+            filter_params.push(format!("%{}%", ea));
+        }
+        if let Some(sym) = &filters.symbol {
+            sql.push_str(" AND symbol = ?");
+            filter_params.push(sym.clone());
+        }
+        if let Some(tf) = &filters.timeframe {
+            sql.push_str(" AND timeframe = ?");
+            filter_params.push(tf.clone());
+        }
+        if let Some(verdict) = &filters.verdict {
+            sql.push_str(" AND verdict = ?");
+            filter_params.push(verdict.clone());
+        }
+
+        // For non-created_at columns, only include rows where that column is not NULL
+        if sort_column != "created_at" {
+            sql.push_str(&format!(" AND {} IS NOT NULL", sort_column));
+        }
+
+        let order = if ascending { "ASC" } else { "DESC" };
+        sql.push_str(&format!(" ORDER BY {} {} LIMIT {}", sort_column, order, limit));
+
+        let mut stmt = conn.prepare(&sql)?;
+        let entries: Vec<ReportEntry> = stmt
+            .query_map(rusqlite::params_from_iter(filter_params.iter()), |row| {
+                let tags_str: String = row.get(23).unwrap_or_else(|_| "[]".to_string());
+                let tags: Vec<String> = serde_json::from_str(&tags_str).unwrap_or_default();
+                Ok(ReportEntry {
+                    id: row.get(0)?,
+                    expert: row.get(1)?,
+                    symbol: row.get(2)?,
+                    timeframe: row.get(3)?,
+                    model: row.get(4)?,
+                    from_date: row.get(5)?,
+                    to_date: row.get(6)?,
+                    created_at: row.get(7)?,
+                    set_file_original: row.get(8)?,
+                    set_snapshot_path: row.get(9)?,
+                    report_dir: row.get(10)?,
+                    charts_dir: row.get(11)?,
+                    net_profit: row.get(12)?,
+                    profit_factor: row.get(13)?,
+                    max_dd_pct: row.get(14)?,
+                    sharpe_ratio: row.get(15)?,
+                    total_trades: row.get(16)?,
+                    win_rate_pct: row.get(17)?,
+                    recovery_factor: row.get(18)?,
+                    deposit: row.get(19)?,
+                    currency: row.get(20)?,
+                    leverage: row.get(21)?,
+                    duration_seconds: row.get(22)?,
+                    tags,
+                    notes: row.get(24)?,
+                    verdict: row.get(25)?,
+                })
+            })?
+            .filter_map(|r| r.ok())
+            .collect();
+
+        Ok(entries)
+    }
+
+    /// Get aggregate statistics for reports
+    pub fn get_stats(&self, filters: &ReportFilters) -> Result<ReportStats> {
+        let conn = self.connect()?;
+
+        let mut sql = "SELECT \
+            COUNT(*), \
+            AVG(net_profit), \
+            AVG(profit_factor), \
+            AVG(max_dd_pct), \
+            AVG(win_rate_pct), \
+            AVG(sharpe_ratio), \
+            SUM(CASE WHEN net_profit > 0 THEN 1 ELSE 0 END), \
+            SUM(CASE WHEN verdict = 'pass' THEN 1 ELSE 0 END), \
+            SUM(CASE WHEN verdict = 'fail' THEN 1 ELSE 0 END), \
+            SUM(CASE WHEN verdict = 'marginal' THEN 1 ELSE 0 END) \
+            FROM reports WHERE 1=1"
+            .to_string();
+
+        let mut filter_params: Vec<String> = Vec::new();
+
+        if let Some(ea) = &filters.expert {
+            sql.push_str(" AND expert LIKE ?");
+            filter_params.push(format!("%{}%", ea));
+        }
+        if let Some(sym) = &filters.symbol {
+            sql.push_str(" AND symbol = ?");
+            filter_params.push(sym.clone());
+        }
+        if let Some(tf) = &filters.timeframe {
+            sql.push_str(" AND timeframe = ?");
+            filter_params.push(tf.clone());
+        }
+        if let Some(verdict) = &filters.verdict {
+            sql.push_str(" AND verdict = ?");
+            filter_params.push(verdict.clone());
+        }
+
+        let stats: ReportStats = conn.query_row(
+            &sql,
+            rusqlite::params_from_iter(filter_params.iter()),
+            |row| {
+                let total: i64 = row.get(0)?;
+                let profitable: Option<i64> = row.get(6)?;
+                let pass_count: Option<i64> = row.get(7)?;
+                let fail_count: Option<i64> = row.get(8)?;
+                let marginal_count: Option<i64> = row.get(9)?;
+
+                Ok(ReportStats {
+                    total_count: total as usize,
+                    avg_net_profit: row.get(1)?,
+                    avg_profit_factor: row.get(2)?,
+                    avg_max_dd_pct: row.get(3)?,
+                    avg_win_rate_pct: row.get(4)?,
+                    avg_sharpe_ratio: row.get(5)?,
+                    profitable_count: profitable.unwrap_or(0) as usize,
+                    pass_verdict_count: pass_count.unwrap_or(0) as usize,
+                    fail_verdict_count: fail_count.unwrap_or(0) as usize,
+                    marginal_verdict_count: marginal_count.unwrap_or(0) as usize,
+                })
+            },
+        )?;
+
+        Ok(stats)
+    }
+}
+
+#[derive(Debug, Serialize, Deserialize, Default)]
+pub struct ReportStats {
+    pub total_count: usize,
+    pub avg_net_profit: Option<f64>,
+    pub avg_profit_factor: Option<f64>,
+    pub avg_max_dd_pct: Option<f64>,
+    pub avg_win_rate_pct: Option<f64>,
+    pub avg_sharpe_ratio: Option<f64>,
+    pub profitable_count: usize,
+    pub pass_verdict_count: usize,
+    pub fail_verdict_count: usize,
+    pub marginal_verdict_count: usize,
 }
@@ -126,3 +126,155 @@ pub fn tool_analyze_concurrent_peak() -> Value {
        }
    })
 }
+
+pub fn tool_list_deals() -> Value {
+    json!({
+        "name": "list_deals",
+        "description": "List individual deals from a backtest report with optional filters",
+        "inputSchema": {
+            "type": "object",
+            "required": ["report_dir"],
+            "properties": {
+                "report_dir": { "type": "string", "description": "Path to report directory containing deals.csv" },
+                "deal_type": { "type": "string", "enum": ["buy", "sell"], "description": "Filter by deal type" },
+                "min_profit": { "type": "number", "description": "Minimum profit (use negative for losses)" },
+                "max_profit": { "type": "number", "description": "Maximum profit" },
+                "start_date": { "type": "string", "description": "Start date filter (YYYY.MM.DD)" },
+                "end_date": { "type": "string", "description": "End date filter (YYYY.MM.DD)" },
+                "min_volume": { "type": "number", "description": "Minimum volume/lots" },
+                "max_volume": { "type": "number", "description": "Maximum volume/lots" },
+                "limit": { "type": "integer", "default": 100, "description": "Max deals to return" }
+            }
+        }
+    })
+}
+
+pub fn tool_search_deals_by_comment() -> Value {
+    json!({
+        "name": "search_deals_by_comment",
+        "description": "Search deals by comment text (case-insensitive partial match)",
+        "inputSchema": {
+            "type": "object",
+            "required": ["report_dir", "query"],
+            "properties": {
+                "report_dir": { "type": "string" },
+                "query": { "type": "string", "description": "Search text in comments" },
+                "limit": { "type": "integer", "default": 50 }
+            }
+        }
+    })
+}
+
+pub fn tool_search_deals_by_magic() -> Value {
+    json!({
+        "name": "search_deals_by_magic",
+        "description": "Filter deals by magic number (EA identifier)",
+        "inputSchema": {
+            "type": "object",
+            "required": ["report_dir", "magic"],
+            "properties": {
+                "report_dir": { "type": "string" },
+                "magic": { "type": "string", "description": "Magic number to filter by" },
+                "limit": { "type": "integer", "default": 100 }
+            }
+        }
+    })
+}
+
+pub fn tool_analyze_profit_distribution() -> Value {
+    json!({
+        "name": "analyze_profit_distribution",
+        "description": "Analyze profit distribution - small/medium/large wins and losses with detailed buckets",
+        "inputSchema": {
+            "type": "object",
+            "required": ["report_dir"],
+            "properties": {
+                "report_dir": { "type": "string" }
+            }
+        }
+    })
+}
+
+pub fn tool_analyze_time_performance() -> Value {
+    json!({
+        "name": "analyze_time_performance",
+        "description": "Analyze performance by hour of day and day of week",
+        "inputSchema": {
+            "type": "object",
+            "required": ["report_dir"],
+            "properties": {
+                "report_dir": { "type": "string" }
+            }
+        }
+    })
+}
+
+pub fn tool_analyze_hold_time_distribution() -> Value {
+    json!({
+        "name": "analyze_hold_time_distribution",
+        "description": "Analyze hold time distribution and correlation with profit",
+        "inputSchema": {
+            "type": "object",
+            "required": ["report_dir"],
+            "properties": {
+                "report_dir": { "type": "string" }
+            }
+        }
+    })
+}
+
+pub fn tool_analyze_layer_performance() -> Value {
+    json!({
+        "name": "analyze_layer_performance",
+        "description": "Analyze performance by grid layer (extracted from deal comments)",
+        "inputSchema": {
+            "type": "object",
+            "required": ["report_dir"],
+            "properties": {
+                "report_dir": { "type": "string" }
+            }
+        }
+    })
+}
+
+pub fn tool_analyze_volume_vs_profit() -> Value {
+    json!({
+        "name": "analyze_volume_vs_profit",
+        "description": "Analyze correlation between volume and profit, plus performance by volume bucket",
+        "inputSchema": {
+            "type": "object",
+            "required": ["report_dir"],
+            "properties": {
+                "report_dir": { "type": "string" }
+            }
+        }
+    })
+}
+
+pub fn tool_analyze_costs() -> Value {
+    json!({
+        "name": "analyze_costs",
+        "description": "Analyze commission and swap costs impact on profitability",
+        "inputSchema": {
+            "type": "object",
+            "required": ["report_dir"],
+            "properties": {
+                "report_dir": { "type": "string" }
+            }
+        }
+    })
+}
+
+pub fn tool_analyze_efficiency() -> Value {
+    json!({
+        "name": "analyze_efficiency",
+        "description": "Calculate efficiency metrics: profit per hour/day, annualized return, trade frequency",
+        "inputSchema": {
+            "type": "object",
+            "required": ["report_dir"],
+            "properties": {
+                "report_dir": { "type": "string" }
+            }
+        }
+    })
+}
@@ -3,7 +3,85 @@ use serde_json::{json, Value};
 pub fn tool_run_backtest() -> Value {
    json!({
        "name": "run_backtest",
-        "description": "Run a complete MT5 backtest pipeline: compile → clean cache → backtest → extract → analyze",
+        "description": "Full backtest pipeline: compile EA → clean cache → run backtest → extract results → analyze. Use this when you have modified the EA source code.",
+        "inputSchema": {
+            "type": "object",
+            "required": ["expert"],
+            "properties": {
+                "expert": { "type": "string", "description": "EA name without path or extension" },
+                "symbol": { "type": "string", "description": "Trading symbol (default: from config or first available)" },
+                "from_date": { "type": "string", "description": "Start date YYYY.MM.DD (default: past complete month)" },
+                "to_date": { "type": "string", "description": "End date YYYY.MM.DD (default: past complete month)" },
+                "timeframe": { "type": "string", "enum": ["M1", "M5", "M15", "M30", "H1", "H4", "D1"], "description": "Chart timeframe (default: M5)" },
+                "deposit": { "type": "integer", "description": "Initial deposit (default: 10000)" },
+                "model": { "type": "integer", "enum": [0, 1, 2], "description": "Tick model: 0=Every tick, 1=OHLC, 2=Open prices" },
+                "set_file": { "type": "string", "description": "Path to .set parameter file for EA inputs" },
+                "skip_compile": { "type": "boolean", "description": "Skip compilation (use existing .ex5)" },
+                "skip_clean": { "type": "boolean", "description": "Skip cache cleaning" },
+                "skip_analyze": { "type": "boolean", "description": "Skip analysis phase" },
+                "deep": { "type": "boolean", "description": "Run deep analysis with extra metrics" },
+                "shutdown": { "type": "boolean", "description": "Close MT5 after backtest completes" },
+                "kill_existing": { "type": "boolean", "description": "Kill any running MT5 instance first" },
+                "timeout": { "type": "integer", "description": "Max wait time in seconds (default: 900)" },
+                "gui": { "type": "boolean", "description": "Enable MT5 visualization window" }
+            }
+        }
+    })
+}
+
+pub fn tool_run_backtest_quick() -> Value {
+    json!({
+        "name": "run_backtest_quick",
+        "description": "Quick backtest using pre-compiled EA: clean cache → run backtest → extract → analyze. Skips compilation. Use when EA code hasn't changed.",
+        "inputSchema": {
+            "type": "object",
+            "required": ["expert"],
+            "properties": {
+                "expert": { "type": "string", "description": "EA name without path or extension (must have .ex5 compiled)" },
+                "symbol": { "type": "string", "description": "Trading symbol (default: from config or first available)" },
+                "from_date": { "type": "string", "description": "Start date YYYY.MM.DD (default: past complete month)" },
+                "to_date": { "type": "string", "description": "End date YYYY.MM.DD (default: past complete month)" },
+                "timeframe": { "type": "string", "enum": ["M1", "M5", "M15", "M30", "H1", "H4", "D1"], "description": "Chart timeframe (default: M5)" },
+                "deposit": { "type": "integer", "description": "Initial deposit (default: 10000)" },
+                "model": { "type": "integer", "enum": [0, 1, 2], "description": "Tick model" },
+                "set_file": { "type": "string", "description": "Path to .set parameter file for EA inputs" },
+                "deep": { "type": "boolean", "description": "Run deep analysis" },
+                "shutdown": { "type": "boolean", "description": "Close MT5 after backtest" },
+                "timeout": { "type": "integer", "description": "Max wait time in seconds (default: 900)" },
+                "gui": { "type": "boolean", "description": "Enable MT5 visualization" }
+            }
+        }
+    })
+}
+
+pub fn tool_run_backtest_only() -> Value {
+    json!({
+        "name": "run_backtest_only",
+        "description": "Backtest only: just run backtest and extract results. No compile, no analysis. Fastest option when you just need raw trade data.",
+        "inputSchema": {
+            "type": "object",
+            "required": ["expert"],
+            "properties": {
+                "expert": { "type": "string", "description": "EA name without path or extension (must have .ex5 compiled)" },
+                "symbol": { "type": "string", "description": "Trading symbol (default: from config)" },
+                "from_date": { "type": "string", "description": "Start date YYYY.MM.DD" },
+                "to_date": { "type": "string", "description": "End date YYYY.MM.DD" },
+                "timeframe": { "type": "string", "enum": ["M1", "M5", "M15", "M30", "H1", "H4", "D1"], "description": "Chart timeframe (default: M5)" },
+                "deposit": { "type": "integer", "description": "Initial deposit (default: 10000)" },
+                "model": { "type": "integer", "enum": [0, 1, 2], "description": "Tick model" },
+                "set_file": { "type": "string", "description": "Path to .set parameter file" },
+                "shutdown": { "type": "boolean", "description": "Close MT5 after backtest" },
+                "timeout": { "type": "integer", "description": "Max wait time (default: 900)" },
+                "gui": { "type": "boolean", "description": "Enable MT5 visualization" }
+            }
+        }
+    })
+}
+
+pub fn tool_launch_backtest() -> Value {
+    json!({
+        "name": "launch_backtest",
+        "description": "Launch MT5 backtest in fire-and-forget mode: compile → clean cache → launch MT5 backtest, then return immediately. Use get_backtest_status to poll for completion.",
        "inputSchema": {
            "type": "object",
            "required": ["expert"],
@@ -18,12 +96,8 @@ pub fn tool_run_backtest() -> Value {
                "set_file": { "type": "string", "description": "Path to .set parameter file" },
                "skip_compile": { "type": "boolean" },
                "skip_clean": { "type": "boolean" },
-                "skip_analyze": { "type": "boolean" },
-                "deep": { "type": "boolean", "description": "Run deep analysis" },
-                "shutdown": { "type": "boolean", "description": "Close MT5 after backtest" },
-                "kill_existing": { "type": "boolean" },
-                "timeout": { "type": "integer" },
-                "gui": { "type": "boolean" }
+                "timeout": { "type": "integer", "description": "Max time in seconds to wait for backtest (default: 900)" },
+                "gui": { "type": "boolean", "description": "Enable visualization during backtest" }
            }
        }
    })
@@ -12,9 +12,12 @@ pub mod utility;

 pub fn get_tools_list() -> Value {
    let tools = vec![
-        // Backtest
-        backtest::tool_run_backtest(),
-        backtest::tool_get_backtest_status(),
+        // Backtest - Granular options
+        backtest::tool_run_backtest(),           // Full pipeline: compile + clean + backtest + extract + analyze
+        backtest::tool_run_backtest_quick(),     // Quick: skip compile, do clean + backtest + extract + analyze
+        backtest::tool_run_backtest_only(),      // Minimal: skip compile, do clean + backtest + extract only
+        backtest::tool_launch_backtest(),        // Fire-and-forget: compile + clean + launch MT5
+        backtest::tool_get_backtest_status(),    // Poll for completion
        backtest::tool_cache_status(),
        backtest::tool_clean_cache(),
        // Optimization
@@ -22,7 +25,7 @@ pub fn get_tools_list() -> Value {
        optimization::tool_get_optimization_status(),
        optimization::tool_get_optimization_results(),
        optimization::tool_list_jobs(),
-        // Analytics (9 tools)
+        // Analytics (19 tools)
        analytics::tool_analyze_report(),
        analytics::tool_analyze_monthly_pnl(),
        analytics::tool_analyze_drawdown_events(),
@@ -32,6 +35,17 @@ pub fn get_tools_list() -> Value {
        analytics::tool_analyze_direction_bias(),
        analytics::tool_analyze_streaks(),
        analytics::tool_analyze_concurrent_peak(),
+        // Deal query tools (10)
+        analytics::tool_list_deals(),
+        analytics::tool_search_deals_by_comment(),
+        analytics::tool_search_deals_by_magic(),
+        analytics::tool_analyze_profit_distribution(),
+        analytics::tool_analyze_time_performance(),
+        analytics::tool_analyze_hold_time_distribution(),
+        analytics::tool_analyze_layer_performance(),
+        analytics::tool_analyze_volume_vs_profit(),
+        analytics::tool_analyze_costs(),
+        analytics::tool_analyze_efficiency(),
        // Baseline
        baseline::tool_compare_baseline(),
        // Experts
@@ -49,6 +63,8 @@ pub fn get_tools_list() -> Value {
        system::tool_list_symbols(),
        system::tool_healthcheck(),
        system::tool_get_active_account(),
+        system::tool_check_update(),
+        system::tool_update(),
        // Utility (8 tools)
        utility::tool_check_symbol_data_status(),
        utility::tool_get_backtest_history(),
@@ -58,6 +74,16 @@ pub fn get_tools_list() -> Value {
        utility::tool_check_mt5_status(),
        utility::tool_create_set_template(),
        utility::tool_export_report(),
+        // Debugging/Diagnostics (10 tools)
+        utility::tool_diagnose_wine(),
+        utility::tool_get_mt5_logs(),
+        utility::tool_search_mt5_errors(),
+        utility::tool_check_mt5_process(),
+        utility::tool_kill_mt5_process(),
+        utility::tool_check_system_resources(),
+        utility::tool_validate_mt5_config(),
+        utility::tool_get_wine_prefix_info(),
+        utility::tool_get_backtest_crash_info(),
        // Set Files
        setfiles::tool_read_set_file(),
        setfiles::tool_write_set_file(),
@@ -67,7 +93,7 @@ pub fn get_tools_list() -> Value {
        setfiles::tool_describe_sweep(),
        setfiles::tool_list_set_files(),
        setfiles::tool_set_from_optimization(),
-        // Reports (11 tools)
+        // Reports (19 tools)
        reports::tool_list_reports(),
        reports::tool_search_reports(),
        reports::tool_get_latest_report(),
@@ -78,6 +104,14 @@ pub fn get_tools_list() -> Value {
        reports::tool_get_history(),
        reports::tool_promote_to_baseline(),
        reports::tool_annotate_history(),
+        reports::tool_get_report_by_id(),
+        reports::tool_get_reports_summary(),
+        reports::tool_get_best_reports(),
+        reports::tool_search_reports_by_tags(),
+        reports::tool_search_reports_by_date_range(),
+        reports::tool_search_reports_by_notes(),
+        reports::tool_get_reports_by_set_file(),
+        reports::tool_get_comparable_reports(),
    ];

    serde_json::json!(tools)
@@ -3,17 +3,17 @@ use serde_json::{json, Value};
 pub fn tool_run_optimization() -> Value {
    json!({
        "name": "run_optimization",
-        "description": "Launch MT5 genetic parameter optimization",
+        "description": "Launch MT5 genetic parameter optimization in fire-and-forget mode. Returns immediately with job_id. Use get_optimization_status to poll for completion. Optimization typically runs for 2-6 hours.",
        "inputSchema": {
            "type": "object",
            "required": ["expert", "set_file", "from_date", "to_date"],
            "properties": {
-                "expert": { "type": "string" },
-                "set_file": { "type": "string" },
-                "symbol": { "type": "string" },
-                "from_date": { "type": "string" },
-                "to_date": { "type": "string" },
-                "deposit": { "type": "integer" }
+                "expert": { "type": "string", "description": "EA name without path or extension" },
+                "set_file": { "type": "string", "description": "Path to .set file with parameter ranges for optimization" },
+                "symbol": { "type": "string", "description": "Trading symbol (default: XAUUSD)" },
+                "from_date": { "type": "string", "description": "Start date YYYY.MM.DD" },
+                "to_date": { "type": "string", "description": "End date YYYY.MM.DD" },
+                "deposit": { "type": "integer", "description": "Initial deposit (default: 10000)" }
            }
        }
    })
@@ -22,12 +22,12 @@ pub fn tool_run_optimization() -> Value {
 pub fn tool_get_optimization_status() -> Value {
    json!({
        "name": "get_optimization_status",
-        "description": "Check progress of a running optimization job",
+        "description": "Check progress of a running optimization job. Poll periodically until status shows 'completed'.",
        "inputSchema": {
            "type": "object",
            "required": ["job_id"],
            "properties": {
-                "job_id": { "type": "string" }
+                "job_id": { "type": "string", "description": "Job ID returned by run_optimization" }
            }
        }
    })
@@ -36,14 +36,14 @@ pub fn tool_get_optimization_status() -> Value {
 pub fn tool_get_optimization_results() -> Value {
    json!({
        "name": "get_optimization_results",
-        "description": "Parse completed MT5 optimization results",
+        "description": "Parse completed MT5 optimization results and find best parameter combinations",
        "inputSchema": {
            "type": "object",
            "properties": {
-                "job_id": { "type": "string" },
-                "report_file": { "type": "string" },
-                "dd_threshold": { "type": "number" },
-                "top_n": { "type": "integer" }
+                "job_id": { "type": "string", "description": "Job ID to parse results for" },
+                "report_file": { "type": "string", "description": "Direct path to optimization report XML file" },
+                "dd_threshold": { "type": "number", "description": "Max drawdown percentage filter" },
+                "top_n": { "type": "integer", "description": "Number of top passes to return (default: 30)" }
            }
        }
    })
@@ -52,7 +52,7 @@ pub fn tool_get_optimization_results() -> Value {
 pub fn tool_list_jobs() -> Value {
    json!({
        "name": "list_jobs",
-        "description": "List running and completed optimization jobs",
+        "description": "List all running and completed optimization jobs with their status",
        "inputSchema": {
            "type": "object"
        }
@@ -148,3 +148,133 @@ pub fn tool_prune_reports() -> Value {
        }
    })
 }
+
+pub fn tool_get_report_by_id() -> Value {
+    json!({
+        "name": "get_report_by_id",
+        "description": "Get a specific report by its ID with full details and optional equity chart",
+        "inputSchema": {
+            "type": "object",
+            "required": ["id"],
+            "properties": {
+                "id": { "type": "string", "description": "Report ID (from list_reports or search_reports)" },
+                "include_chart": { "type": "boolean", "description": "Include equity chart as base64 PNG (default: true)", "default": true }
+            }
+        }
+    })
+}
+
+pub fn tool_get_reports_summary() -> Value {
+    json!({
+        "name": "get_reports_summary",
+        "description": "Get aggregate statistics across reports - counts, averages by EA/symbol/timeframe/verdict",
+        "inputSchema": {
+            "type": "object",
+            "properties": {
+                "expert": { "type": "string", "description": "Filter by EA name substring" },
+                "symbol": { "type": "string", "description": "Filter by exact symbol" },
+                "timeframe": { "type": "string", "description": "Filter by exact timeframe" },
+                "verdict": { "type": "string", "description": "Filter by verdict (pass/fail/marginal)" }
+            }
+        }
+    })
+}
+
+pub fn tool_get_best_reports() -> Value {
+    json!({
+        "name": "get_best_reports",
+        "description": "Get top N reports sorted by performance metric (profit factor, win rate, drawdown, etc.)",
+        "inputSchema": {
+            "type": "object",
+            "properties": {
+                "sort_by": { "type": "string", "enum": ["net_profit", "profit_factor", "max_dd_pct", "win_rate_pct", "sharpe_ratio", "recovery_factor", "total_trades"], "default": "profit_factor", "description": "Metric to sort by" },
+                "order": { "type": "string", "enum": ["asc", "desc"], "default": "desc", "description": "Sort order (use 'asc' for drawdown, 'desc' for profit)" },
+                "limit": { "type": "integer", "default": 10, "description": "Number of reports to return" },
+                "expert": { "type": "string", "description": "Filter by EA name substring" },
+                "symbol": { "type": "string", "description": "Filter by exact symbol" },
+                "timeframe": { "type": "string", "description": "Filter by exact timeframe" },
+                "verdict": { "type": "string", "description": "Filter by verdict" }
+            }
+        }
+    })
+}
+
+pub fn tool_search_reports_by_tags() -> Value {
+    json!({
+        "name": "search_reports_by_tags",
+        "description": "Search reports by tags - at least one tag must match (OR logic)",
+        "inputSchema": {
+            "type": "object",
+            "required": ["tags"],
+            "properties": {
+                "tags": { "type": "array", "items": { "type": "string" }, "description": "Tags to search for (e.g., ['production', 'verified'])" },
+                "limit": { "type": "integer", "default": 50 }
+            }
+        }
+    })
+}
+
+pub fn tool_search_reports_by_date_range() -> Value {
+    json!({
+        "name": "search_reports_by_date_range",
+        "description": "Search reports by backtest date range (from_date and to_date fields)",
+        "inputSchema": {
+            "type": "object",
+            "properties": {
+                "from_start": { "type": "string", "description": "From date >= this (YYYY.MM.DD)" },
+                "from_end": { "type": "string", "description": "From date <= this (YYYY.MM.DD)" },
+                "to_start": { "type": "string", "description": "To date >= this (YYYY.MM.DD)" },
+                "to_end": { "type": "string", "description": "To date <= this (YYYY.MM.DD)" },
+                "limit": { "type": "integer", "default": 50 }
+            }
+        }
+    })
+}
+
+pub fn tool_search_reports_by_notes() -> Value {
+    json!({
+        "name": "search_reports_by_notes",
+        "description": "Full-text search in report notes field (case-insensitive LIKE search)",
+        "inputSchema": {
+            "type": "object",
+            "required": ["query"],
+            "properties": {
+                "query": { "type": "string", "description": "Search text (partial match)" },
+                "limit": { "type": "integer", "default": 50 }
+            }
+        }
+    })
+}
+
+pub fn tool_get_reports_by_set_file() -> Value {
+    json!({
+        "name": "get_reports_by_set_file",
+        "description": "Find all reports that used a specific .set parameter file",
+        "inputSchema": {
+            "type": "object",
+            "required": ["set_file"],
+            "properties": {
+                "set_file": { "type": "string", "description": "Set filename or partial path to match" },
+                "limit": { "type": "integer", "default": 50 }
+            }
+        }
+    })
+}
+
+pub fn tool_get_comparable_reports() -> Value {
+    json!({
+        "name": "get_comparable_reports",
+        "description": "Find reports comparable to a given report (same EA, symbol, timeframe) - useful for before/after analysis",
+        "inputSchema": {
+            "type": "object",
+            "properties": {
+                "report_id": { "type": "string", "description": "Reference report ID (if provided, uses its expert/symbol/timeframe)" },
+                "expert": { "type": "string", "description": "EA name (required if report_id not provided)" },
+                "symbol": { "type": "string", "description": "Symbol (required if report_id not provided)" },
+                "timeframe": { "type": "string", "description": "Timeframe (required if report_id not provided)" },
+                "exclude_id": { "type": "string", "description": "Exclude this report ID from results" },
+                "limit": { "type": "integer", "default": 20 }
+            }
+        }
+    })
+}
@@ -45,3 +45,25 @@ pub fn tool_get_active_account() -> Value {
        }
    })
 }
+
+pub fn tool_check_update() -> Value {
+    json!({
+        "name": "check_update",
+        "description": "Check if a newer version of mt5-quant is available on GitHub. A background check runs automatically on the first tool call of each session; this tool returns that cached result instantly or fetches it on demand.",
+        "inputSchema": {
+            "type": "object",
+            "properties": {}
+        }
+    })
+}
+
+pub fn tool_update() -> Value {
+    json!({
+        "name": "update",
+        "description": "Download and install the latest mt5-quant binary from GitHub Releases, then replace the current executable in place. Restart the MCP connection after updating to load the new version.",
+        "inputSchema": {
+            "type": "object",
+            "properties": {}
+        }
+    })
+}
@@ -165,3 +165,149 @@ pub fn tool_export_report() -> Value {
        }
    })
 }
+
+pub fn tool_diagnose_wine() -> Value {
+    json!({
+        "name": "diagnose_wine",
+        "description": "Check Wine installation, version, and prefix health. Reports errors, warnings, and recent Wine errors.",
+        "inputSchema": {
+            "type": "object"
+        }
+    })
+}
+
+pub fn tool_get_mt5_logs() -> Value {
+    json!({
+        "name": "get_mt5_logs",
+        "description": "Get MT5 terminal, tester, or MetaEditor logs with optional search filtering",
+        "inputSchema": {
+            "type": "object",
+            "properties": {
+                "log_type": {
+                    "type": "string",
+                    "enum": ["terminal", "tester", "metaeditor"],
+                    "default": "terminal",
+                    "description": "Type of log to retrieve"
+                },
+                "lines": {
+                    "type": "integer",
+                    "default": 100,
+                    "description": "Number of lines to return (from end of log)"
+                },
+                "search": {
+                    "type": "string",
+                    "description": "Optional search term to filter log lines"
+                }
+            }
+        }
+    })
+}
+
+pub fn tool_search_mt5_errors() -> Value {
+    json!({
+        "name": "search_mt5_errors",
+        "description": "Search MT5 logs for error patterns (error, failed, crash, exception, etc.) in recent hours",
+        "inputSchema": {
+            "type": "object",
+            "properties": {
+                "hours_back": {
+                    "type": "integer",
+                    "default": 24,
+                    "description": "Hours to search back in logs"
+                },
+                "max_errors": {
+                    "type": "integer",
+                    "default": 50,
+                    "description": "Maximum number of errors to return"
+                }
+            }
+        }
+    })
+}
+
+pub fn tool_check_mt5_process() -> Value {
+    json!({
+        "name": "check_mt5_process",
+        "description": "Check if MT5 processes are running, get process info (PID, CPU, memory usage)",
+        "inputSchema": {
+            "type": "object"
+        }
+    })
+}
+
+pub fn tool_kill_mt5_process() -> Value {
+    json!({
+        "name": "kill_mt5_process",
+        "description": "Kill stuck MT5 processes. Use force=true for stuck wineserver.",
+        "inputSchema": {
+            "type": "object",
+            "properties": {
+                "pid": {
+                    "type": "string",
+                    "description": "Optional specific PID to kill"
+                },
+                "force": {
+                    "type": "boolean",
+                    "default": false,
+                    "description": "Use SIGKILL (-9) instead of SIGTERM (-15), also kills wineserver"
+                }
+            }
+        }
+    })
+}
+
+pub fn tool_check_system_resources() -> Value {
+    json!({
+        "name": "check_system_resources",
+        "description": "Check disk space, memory, and CPU. Warns if resources are low for MT5 operations.",
+        "inputSchema": {
+            "type": "object"
+        }
+    })
+}
+
+pub fn tool_validate_mt5_config() -> Value {
+    json!({
+        "name": "validate_mt5_config",
+        "description": "Validate MT5 configuration files (terminal.ini, tester settings). Reports errors and warnings.",
+        "inputSchema": {
+            "type": "object"
+        }
+    })
+}
+
+pub fn tool_get_wine_prefix_info() -> Value {
+    json!({
+        "name": "get_wine_prefix_info",
+        "description": "Get detailed Wine prefix information: Windows version, installed programs, registry files, drive_c size",
+        "inputSchema": {
+            "type": "object"
+        }
+    })
+}
+
+pub fn tool_get_backtest_crash_info() -> Value {
+    json!({
+        "name": "get_backtest_crash_info",
+        "description": "Investigate backtest crashes/failures. Checks for incomplete markers, missing deals.csv, error logs. Can scan recent reports.",
+        "inputSchema": {
+            "type": "object",
+            "properties": {
+                "report_dir": {
+                    "type": "string",
+                    "description": "Optional specific report directory to check"
+                },
+                "check_recent": {
+                    "type": "boolean",
+                    "default": true,
+                    "description": "Also check recent reports for failures"
+                },
+                "hours_back": {
+                    "type": "integer",
+                    "default": 6,
+                    "description": "Hours back to check for recent failures"
+                }
+            }
+        }
+    })
+}
@@ -6,8 +6,30 @@ use std::path::Path;
 use crate::analytics::DealAnalyzer;
 use crate::models::deals::Deal;
 use crate::models::metrics::Metrics;
+use crate::models::Config;
+
+// ── Internal helpers ──────────────────────────────────────────────────────────
+
+fn required_str<'a>(args: &'a Value, key: &str) -> Result<&'a str> {
+    args.get(key)
+        .and_then(|v| v.as_str())
+        .ok_or_else(|| anyhow::anyhow!("{} is required", key))
+}
+
+fn ok_response(data: Value) -> Value {
+    json!({
+        "content": [{ "type": "text", "text": data.to_string() }],
+        "isError": false
+    })
+}
+
+fn err_response(msg: impl std::fmt::Display) -> Value {
+    json!({
+        "content": [{ "type": "text", "text": msg.to_string() }],
+        "isError": true
+    })
+}

-/// Helper to load deals and metrics from report directory
 fn load_report_data(report_dir: &str) -> Result<(Vec<Deal>, Metrics)> {
    let deals_csv = Path::new(report_dir).join("deals.csv");
    let metrics_json = Path::new(report_dir).join("metrics.json");
@@ -17,10 +39,8 @@ fn load_report_data(report_dir: &str) -> Result<(Vec<Deal>, Metrics)> {
    }

    let deals = read_deals_from_csv(&deals_csv)?;
-    
    let metrics = if metrics_json.exists() {
-        let content = fs::read_to_string(&metrics_json)?;
-        serde_json::from_str(&content)?
+        serde_json::from_str(&fs::read_to_string(&metrics_json)?)?
    } else {
        Metrics::default()
    };
@@ -28,75 +48,18 @@ fn load_report_data(report_dir: &str) -> Result<(Vec<Deal>, Metrics)> {
    Ok((deals, metrics))
 }

-pub async fn handle_analyze_report(_config: &Config, args: &Value) -> Result<Value> {
-    let report_dir = args.get("report_dir")
-        .and_then(|v| v.as_str())
-        .ok_or_else(|| anyhow::anyhow!("report_dir is required"))?;
-
+fn prepare_analysis(report_dir: &str) -> Result<(Vec<Deal>, Metrics, DealAnalyzer)> {
    let (deals, metrics) = load_report_data(report_dir)?;
-    let analyzer = DealAnalyzer::new();
-
-    // Check if specific analytics requested
-    let requested: Option<HashSet<String>> = args.get("analytics")
-        .and_then(|v| v.as_array())
-        .map(|arr| arr.iter()
-            .filter_map(|v| v.as_str().map(|s| s.to_string()))
-            .collect());
-
-    let top_losses_limit = args.get("top_losses_limit").and_then(|v| v.as_u64()).unwrap_or(10) as usize;
-
-    let all = requested.is_none();
-    let req = |name: &str| all || requested.as_ref().map(|s| s.contains(name)).unwrap_or(false);
-
-    // Build selective result
-    let mut result = json!({});
-
-    if req("monthly_pnl") || all {
-        result["monthly"] = json!(analyzer.monthly_pnl(&deals));
-    }
-    if req("drawdown_events") || all {
-        result["dd_events"] = json!(analyzer.reconstruct_dd_events(&deals, &metrics));
-    }
-    if req("top_losses") || all {
-        result["top_losses"] = json!(analyzer.top_losses(&deals, top_losses_limit));
-    }
-    if req("loss_sequences") || all {
-        result["loss_sequences"] = json!(analyzer.loss_sequences(&deals));
-    }
-    if req("position_pairs") || all {
-        result["position_pairs"] = json!(analyzer.position_pairs(&deals));
-    }
-    if req("direction_bias") || all {
-        result["direction_bias"] = json!(analyzer.direction_bias(&deals));
-    }
-    if req("streak_analysis") || all {
-        result["streak_analysis"] = json!(analyzer.streak_analysis(&deals));
-    }
-    if req("concurrent_peak") || all {
-        result["concurrent_peak"] = json!(analyzer.concurrent_peak(&deals));
-    }
-
-    let analysis_path = Path::new(report_dir).join("analysis.json");
-    fs::write(&analysis_path, serde_json::to_string_pretty(&result)?)?;
-
-    Ok(json!({
-        "content": [{ "type": "text", "text": json!({
-            "success": true,
-            "analysis_file": analysis_path.to_string_lossy(),
-            "analytics_run": requested.map(|s| s.iter().cloned().collect::<Vec<_>>()).unwrap_or_else(|| vec!["all".to_string()]),
-            "summary": result,
-        }).to_string() }],
-        "isError": false
-    }))
+    Ok((deals, metrics, DealAnalyzer::new()))
 }

 fn read_deals_from_csv(path: &Path) -> Result<Vec<Deal>> {
    let content = fs::read_to_string(path)?;
    let mut deals = Vec::new();
-    
+
    let mut lines = content.lines();
    let _header = lines.next();
-    
+
    for line in lines {
        let parts: Vec<&str> = line.split(',').collect();
        if parts.len() >= 12 {
@@ -118,192 +81,268 @@ fn read_deals_from_csv(path: &Path) -> Result<Vec<Deal>> {
            });
        }
    }
-    
+
    Ok(deals)
 }

+// ── Composite analytics ───────────────────────────────────────────────────────
+
+pub async fn handle_analyze_report(_config: &Config, args: &Value) -> Result<Value> {
+    let report_dir = required_str(args, "report_dir")?;
+    let (deals, metrics, analyzer) = prepare_analysis(report_dir)?;
+
+    let requested: Option<HashSet<String>> = args.get("analytics")
+        .and_then(|v| v.as_array())
+        .map(|arr| arr.iter().filter_map(|v| v.as_str().map(|s| s.to_string())).collect());
+
+    let top_losses_limit = args.get("top_losses_limit").and_then(|v| v.as_u64()).unwrap_or(10) as usize;
+
+    let all = requested.is_none();
+    let req = |name: &str| all || requested.as_ref().map(|s| s.contains(name)).unwrap_or(false);
+
+    let mut result = json!({});
+
+    if req("monthly_pnl")      { result["monthly"]          = json!(analyzer.monthly_pnl(&deals)); }
+    if req("drawdown_events")  { result["dd_events"]        = json!(analyzer.reconstruct_dd_events(&deals, &metrics)); }
+    if req("top_losses")       { result["top_losses"]       = json!(analyzer.top_losses(&deals, top_losses_limit)); }
+    if req("loss_sequences")   { result["loss_sequences"]   = json!(analyzer.loss_sequences(&deals)); }
+    if req("position_pairs")   { result["position_pairs"]   = json!(analyzer.position_pairs(&deals)); }
+    if req("direction_bias")   { result["direction_bias"]   = json!(analyzer.direction_bias(&deals)); }
+    if req("streak_analysis")  { result["streak_analysis"]  = json!(analyzer.streak_analysis(&deals)); }
+    if req("concurrent_peak")  { result["concurrent_peak"]  = json!(analyzer.concurrent_peak(&deals)); }
+
+    let analysis_path = Path::new(report_dir).join("analysis.json");
+    fs::write(&analysis_path, serde_json::to_string_pretty(&result)?)?;
+
+    Ok(ok_response(json!({
+        "success": true,
+        "analysis_file": analysis_path.to_string_lossy(),
+        "analytics_run": requested.map(|s| s.iter().cloned().collect::<Vec<_>>()).unwrap_or_else(|| vec!["all".to_string()]),
+        "summary": result,
+    })))
+}
+
 pub async fn handle_compare_baseline(_config: &Config, args: &Value) -> Result<Value> {
-    let report_dir = args.get("report_dir")
-        .and_then(|v| v.as_str())
-        .ok_or_else(|| anyhow::anyhow!("report_dir is required"))?;
+    let report_dir = required_str(args, "report_dir")?;

    let baseline_path = Path::new("config/baseline.json");
    let metrics_path = Path::new(report_dir).join("metrics.json");

    if !baseline_path.exists() {
-        return Ok(json!({
-            "content": [{ "type": "text", "text": "No baseline.json found in config/" }],
-            "isError": false
-        }));
+        return Ok(ok_response(json!("No baseline.json found in config/")));
    }

    let baseline: Value = serde_json::from_str(&fs::read_to_string(baseline_path)?)?;
    let current: Value = serde_json::from_str(&fs::read_to_string(metrics_path)?)?;

-    let comparison = json!({
+    Ok(ok_response(json!({
        "baseline": baseline,
        "current": current,
        "improvements": {
-            "profit": current.get("net_profit").and_then(|v| v.as_f64()).unwrap_or(0.0) 
+            "profit": current.get("net_profit").and_then(|v| v.as_f64()).unwrap_or(0.0)
                - baseline.get("net_profit").and_then(|v| v.as_f64()).unwrap_or(0.0),
            "drawdown": current.get("max_dd_pct").and_then(|v| v.as_f64()).unwrap_or(0.0)
                - baseline.get("max_dd_pct").and_then(|v| v.as_f64()).unwrap_or(0.0),
        }
-    });
-
-    Ok(json!({
-        "content": [{ "type": "text", "text": comparison.to_string() }],
-        "isError": false
-    }))
+    })))
 }

-// === Granular Analytics Handlers ===
+// ── Granular analytics handlers ───────────────────────────────────────────────

 pub async fn handle_analyze_monthly_pnl(_config: &Config, args: &Value) -> Result<Value> {
-    let report_dir = args.get("report_dir")
-        .and_then(|v| v.as_str())
-        .ok_or_else(|| anyhow::anyhow!("report_dir is required"))?;
-
-    let (deals, _) = load_report_data(report_dir)?;
-    let analyzer = DealAnalyzer::new();
-    let result = analyzer.monthly_pnl(&deals);
-
-    Ok(json!({
-        "content": [{ "type": "text", "text": json!({
-            "success": true,
-            "monthly_pnl": result,
-        }).to_string() }],
-        "isError": false
-    }))
+    let report_dir = required_str(args, "report_dir")?;
+    let (deals, _, analyzer) = prepare_analysis(report_dir)?;
+    Ok(ok_response(json!({ "success": true, "monthly_pnl": analyzer.monthly_pnl(&deals) })))
 }

 pub async fn handle_analyze_drawdown_events(_config: &Config, args: &Value) -> Result<Value> {
-    let report_dir = args.get("report_dir")
-        .and_then(|v| v.as_str())
-        .ok_or_else(|| anyhow::anyhow!("report_dir is required"))?;
-
-    let (deals, metrics) = load_report_data(report_dir)?;
-    let analyzer = DealAnalyzer::new();
-    let result = analyzer.reconstruct_dd_events(&deals, &metrics);
-
-    Ok(json!({
-        "content": [{ "type": "text", "text": json!({
-            "success": true,
-            "drawdown_events": result,
-        }).to_string() }],
-        "isError": false
-    }))
+    let report_dir = required_str(args, "report_dir")?;
+    let (deals, metrics, analyzer) = prepare_analysis(report_dir)?;
+    Ok(ok_response(json!({ "success": true, "drawdown_events": analyzer.reconstruct_dd_events(&deals, &metrics) })))
 }

 pub async fn handle_analyze_top_losses(_config: &Config, args: &Value) -> Result<Value> {
-    let report_dir = args.get("report_dir")
-        .and_then(|v| v.as_str())
-        .ok_or_else(|| anyhow::anyhow!("report_dir is required"))?;
-
+    let report_dir = required_str(args, "report_dir")?;
    let limit = args.get("limit").and_then(|v| v.as_u64()).unwrap_or(10) as usize;
-    let (deals, _) = load_report_data(report_dir)?;
-    let analyzer = DealAnalyzer::new();
-    let result = analyzer.top_losses(&deals, limit);
-
-    Ok(json!({
-        "content": [{ "type": "text", "text": json!({
-            "success": true,
-            "limit": limit,
-            "top_losses": result,
-        }).to_string() }],
-        "isError": false
-    }))
+    let (deals, _, analyzer) = prepare_analysis(report_dir)?;
+    Ok(ok_response(json!({ "success": true, "limit": limit, "top_losses": analyzer.top_losses(&deals, limit) })))
 }

 pub async fn handle_analyze_loss_sequences(_config: &Config, args: &Value) -> Result<Value> {
-    let report_dir = args.get("report_dir")
-        .and_then(|v| v.as_str())
-        .ok_or_else(|| anyhow::anyhow!("report_dir is required"))?;
-
-    let (deals, _) = load_report_data(report_dir)?;
-    let analyzer = DealAnalyzer::new();
-    let result = analyzer.loss_sequences(&deals);
-
-    Ok(json!({
-        "content": [{ "type": "text", "text": json!({
-            "success": true,
-            "loss_sequences": result,
-        }).to_string() }],
-        "isError": false
-    }))
+    let report_dir = required_str(args, "report_dir")?;
+    let (deals, _, analyzer) = prepare_analysis(report_dir)?;
+    Ok(ok_response(json!({ "success": true, "loss_sequences": analyzer.loss_sequences(&deals) })))
 }

 pub async fn handle_analyze_position_pairs(_config: &Config, args: &Value) -> Result<Value> {
-    let report_dir = args.get("report_dir")
-        .and_then(|v| v.as_str())
-        .ok_or_else(|| anyhow::anyhow!("report_dir is required"))?;
-
-    let (deals, _) = load_report_data(report_dir)?;
-    let analyzer = DealAnalyzer::new();
-    let result = analyzer.position_pairs(&deals);
-
-    Ok(json!({
-        "content": [{ "type": "text", "text": json!({
-            "success": true,
-            "position_pairs": result,
-        }).to_string() }],
-        "isError": false
-    }))
+    let report_dir = required_str(args, "report_dir")?;
+    let (deals, _, analyzer) = prepare_analysis(report_dir)?;
+    Ok(ok_response(json!({ "success": true, "position_pairs": analyzer.position_pairs(&deals) })))
 }

 pub async fn handle_analyze_direction_bias(_config: &Config, args: &Value) -> Result<Value> {
-    let report_dir = args.get("report_dir")
-        .and_then(|v| v.as_str())
-        .ok_or_else(|| anyhow::anyhow!("report_dir is required"))?;
-
-    let (deals, _) = load_report_data(report_dir)?;
-    let analyzer = DealAnalyzer::new();
-    let result = analyzer.direction_bias(&deals);
-
-    Ok(json!({
-        "content": [{ "type": "text", "text": json!({
-            "success": true,
-            "direction_bias": result,
-        }).to_string() }],
-        "isError": false
-    }))
+    let report_dir = required_str(args, "report_dir")?;
+    let (deals, _, analyzer) = prepare_analysis(report_dir)?;
+    Ok(ok_response(json!({ "success": true, "direction_bias": analyzer.direction_bias(&deals) })))
 }

 pub async fn handle_analyze_streaks(_config: &Config, args: &Value) -> Result<Value> {
-    let report_dir = args.get("report_dir")
-        .and_then(|v| v.as_str())
-        .ok_or_else(|| anyhow::anyhow!("report_dir is required"))?;
-
-    let (deals, _) = load_report_data(report_dir)?;
-    let analyzer = DealAnalyzer::new();
-    let result = analyzer.streak_analysis(&deals);
-
-    Ok(json!({
-        "content": [{ "type": "text", "text": json!({
-            "success": true,
-            "streak_analysis": result,
-        }).to_string() }],
-        "isError": false
-    }))
+    let report_dir = required_str(args, "report_dir")?;
+    let (deals, _, analyzer) = prepare_analysis(report_dir)?;
+    Ok(ok_response(json!({ "success": true, "streak_analysis": analyzer.streak_analysis(&deals) })))
 }

 pub async fn handle_analyze_concurrent_peak(_config: &Config, args: &Value) -> Result<Value> {
-    let report_dir = args.get("report_dir")
-        .and_then(|v| v.as_str())
-        .ok_or_else(|| anyhow::anyhow!("report_dir is required"))?;
-
-    let (deals, _) = load_report_data(report_dir)?;
-    let analyzer = DealAnalyzer::new();
-    let result = analyzer.concurrent_peak(&deals);
-
-    Ok(json!({
-        "content": [{ "type": "text", "text": json!({
-            "success": true,
-            "concurrent_peak": result,
-        }).to_string() }],
-        "isError": false
-    }))
+    let report_dir = required_str(args, "report_dir")?;
+    let (deals, _, analyzer) = prepare_analysis(report_dir)?;
+    Ok(ok_response(json!({ "success": true, "concurrent_peak": analyzer.concurrent_peak(&deals) })))
 }

-// Import Config for analysis module
-use crate::models::Config;
+pub async fn handle_analyze_profit_distribution(_config: &Config, args: &Value) -> Result<Value> {
+    let report_dir = required_str(args, "report_dir")?;
+    let (deals, _, analyzer) = prepare_analysis(report_dir)?;
+    Ok(ok_response(json!({ "success": true, "profit_distribution": analyzer.profit_distribution(&deals) })))
+}
+
+pub async fn handle_analyze_time_performance(_config: &Config, args: &Value) -> Result<Value> {
+    let report_dir = required_str(args, "report_dir")?;
+    let (deals, _, analyzer) = prepare_analysis(report_dir)?;
+    Ok(ok_response(json!({ "success": true, "time_performance": analyzer.time_performance(&deals) })))
+}
+
+pub async fn handle_analyze_hold_time_distribution(_config: &Config, args: &Value) -> Result<Value> {
+    let report_dir = required_str(args, "report_dir")?;
+    let (deals, _, analyzer) = prepare_analysis(report_dir)?;
+    Ok(ok_response(json!({ "success": true, "hold_time_analysis": analyzer.hold_time_analysis(&deals) })))
+}
+
+pub async fn handle_analyze_layer_performance(_config: &Config, args: &Value) -> Result<Value> {
+    let report_dir = required_str(args, "report_dir")?;
+    let (deals, _, analyzer) = prepare_analysis(report_dir)?;
+    Ok(ok_response(json!({ "success": true, "layer_performance": analyzer.layer_performance(&deals) })))
+}
+
+pub async fn handle_analyze_volume_vs_profit(_config: &Config, args: &Value) -> Result<Value> {
+    let report_dir = required_str(args, "report_dir")?;
+    let (deals, _, analyzer) = prepare_analysis(report_dir)?;
+    Ok(ok_response(json!({ "success": true, "volume_analysis": analyzer.volume_analysis(&deals) })))
+}
+
+pub async fn handle_analyze_costs(_config: &Config, args: &Value) -> Result<Value> {
+    let report_dir = required_str(args, "report_dir")?;
+    let (deals, _, analyzer) = prepare_analysis(report_dir)?;
+    Ok(ok_response(json!({ "success": true, "cost_analysis": analyzer.cost_analysis(&deals) })))
+}
+
+pub async fn handle_analyze_efficiency(_config: &Config, args: &Value) -> Result<Value> {
+    let report_dir = required_str(args, "report_dir")?;
+    let (deals, metrics, analyzer) = prepare_analysis(report_dir)?;
+    Ok(ok_response(json!({ "success": true, "efficiency_analysis": analyzer.efficiency_analysis(&deals, &metrics) })))
+}
+
+// ── Deal query handlers ───────────────────────────────────────────────────────
+
+fn deal_to_json(d: &Deal) -> Value {
+    json!({
+        "time": d.time,
+        "deal": d.deal,
+        "symbol": d.symbol,
+        "deal_type": d.deal_type,
+        "volume": d.volume,
+        "price": d.price,
+        "profit": d.profit,
+        "commission": d.commission,
+        "swap": d.swap,
+        "comment": d.comment,
+        "magic": d.magic,
+    })
+}
+
+fn is_closed_trade(d: &Deal) -> bool {
+    d.entry.to_lowercase().contains("out") && d.profit != 0.0
+}
+
+pub async fn handle_list_deals(_config: &Config, args: &Value) -> Result<Value> {
+    let report_dir = required_str(args, "report_dir")?;
+    let (deals, _) = load_report_data(report_dir)?;
+
+    let deal_type  = args.get("deal_type").and_then(|v| v.as_str());
+    let min_profit = args.get("min_profit").and_then(|v| v.as_f64());
+    let max_profit = args.get("max_profit").and_then(|v| v.as_f64());
+    let start_date = args.get("start_date").and_then(|v| v.as_str());
+    let end_date   = args.get("end_date").and_then(|v| v.as_str());
+    let min_volume = args.get("min_volume").and_then(|v| v.as_f64());
+    let max_volume = args.get("max_volume").and_then(|v| v.as_f64());
+    let limit = args.get("limit").and_then(|v| v.as_u64()).unwrap_or(100) as usize;
+
+    let mut filtered: Vec<&Deal> = deals.iter().filter(|d| {
+        if !is_closed_trade(d) { return false; }
+        if let Some(dt)  = deal_type  { if !d.deal_type.to_lowercase().contains(dt)           { return false; } }
+        if let Some(min) = min_profit { if d.profit < min                                      { return false; } }
+        if let Some(max) = max_profit { if d.profit > max                                      { return false; } }
+        if let Some(s)   = start_date { if d.time.as_str() < s                                 { return false; } }
+        if let Some(e)   = end_date   { if d.time.as_str() > e                                 { return false; } }
+        if let Some(min) = min_volume { if d.volume < min                                      { return false; } }
+        if let Some(max) = max_volume { if d.volume > max                                      { return false; } }
+        true
+    }).collect();
+
+    filtered.sort_by(|a, b| b.time.cmp(&a.time));
+    filtered.truncate(limit);
+
+    Ok(ok_response(json!({
+        "success": true,
+        "total_deals": deals.len(),
+        "filtered_count": filtered.len(),
+        "deals": filtered.iter().map(|d| deal_to_json(d)).collect::<Vec<_>>(),
+    })))
+}
+
+pub async fn handle_search_deals_by_comment(_config: &Config, args: &Value) -> Result<Value> {
+    let report_dir = required_str(args, "report_dir")?;
+    let query = required_str(args, "query")?;
+    let limit = args.get("limit").and_then(|v| v.as_u64()).unwrap_or(50) as usize;
+
+    let (deals, _) = load_report_data(report_dir)?;
+    let query_lower = query.to_lowercase();
+
+    let mut filtered: Vec<&Deal> = deals.iter()
+        .filter(|d| is_closed_trade(d) && d.comment.to_lowercase().contains(&query_lower))
+        .collect();
+
+    filtered.sort_by(|a, b| b.time.cmp(&a.time));
+    filtered.truncate(limit);
+
+    Ok(ok_response(json!({
+        "success": true,
+        "query": query,
+        "matched": filtered.len(),
+        "deals": filtered.iter().map(|d| deal_to_json(d)).collect::<Vec<_>>(),
+    })))
+}
+
+pub async fn handle_search_deals_by_magic(_config: &Config, args: &Value) -> Result<Value> {
+    let report_dir = required_str(args, "report_dir")?;
+    let magic = required_str(args, "magic")?;
+    let limit = args.get("limit").and_then(|v| v.as_u64()).unwrap_or(100) as usize;
+
+    let (deals, _) = load_report_data(report_dir)?;
+
+    let mut filtered: Vec<&Deal> = deals.iter()
+        .filter(|d| is_closed_trade(d) && d.magic.as_ref().map(|m| m.contains(magic)).unwrap_or(false))
+        .collect();
+
+    filtered.sort_by(|a, b| b.time.cmp(&a.time));
+    filtered.truncate(limit);
+
+    Ok(ok_response(json!({
+        "success": true,
+        "magic": magic,
+        "matched": filtered.len(),
+        "deals": filtered.iter().map(|d| deal_to_json(d)).collect::<Vec<_>>(),
+    })))
+}
+
+// suppress unused warning — err_response is available for future handlers
+#[allow(dead_code)]
+fn _use_err_response() { let _ = err_response(""); }
@@ -2,7 +2,9 @@ use anyhow::Result;
 use serde_json::{json, Value};
 use std::fs;
 use std::path::Path;
+use std::process::Command;
 use crate::models::Config;
+use crate::models::report::BacktestJob;
 use crate::pipeline::backtest::{BacktestParams, BacktestPipeline};

 /// Pre-flight check result for backtest readiness
@@ -190,38 +192,243 @@ pub async fn handle_run_backtest(config: &Config, args: &Value) -> Result<Value>
    }))
 }

+pub async fn handle_run_backtest_quick(config: &Config, args: &Value) -> Result<Value> {
+    // Quick backtest: skip compile, do clean → backtest → extract → analyze
+    let mut args = args.clone();
+    if let Some(obj) = args.as_object_mut() {
+        obj.insert("skip_compile".to_string(), json!(true));
+        // keep skip_analyze as false (default) to run analysis
+    }
+    handle_run_backtest(config, &args).await
+}
+
+pub async fn handle_run_backtest_only(config: &Config, args: &Value) -> Result<Value> {
+    // Backtest only: skip compile, skip analyze - just backtest and extract
+    let mut args = args.clone();
+    if let Some(obj) = args.as_object_mut() {
+        obj.insert("skip_compile".to_string(), json!(true));
+        obj.insert("skip_analyze".to_string(), json!(true));
+    }
+    handle_run_backtest(config, &args).await
+}
+
+pub async fn handle_launch_backtest(config: &Config, args: &Value) -> Result<Value> {
+    let expert = args.get("expert")
+        .and_then(|v| v.as_str())
+        .ok_or_else(|| anyhow::anyhow!("expert is required"))?;
+
+    // Run pre-flight check
+    let preflight = BacktestPreflight::check(config, expert);
+    
+    // Check account session
+    if preflight.account.is_none() {
+        return Ok(json!({
+            "content": [{ "type": "text", "text": json!({
+                "error": "No active MT5 account session detected.",
+                "hint": "Open MT5 and login to your trading account before running backtests."
+            }).to_string() }],
+            "isError": true
+        }));
+    }
+    
+    // Get symbol
+    let requested_symbol = args.get("symbol")
+        .and_then(|v| v.as_str())
+        .unwrap_or("");
+
+    let symbol = if requested_symbol.is_empty() {
+        config.backtest_symbol.clone()
+            .or_else(|| preflight.available_symbols.first().cloned())
+            .unwrap_or_else(|| "EURUSD".to_string())
+    } else {
+        requested_symbol.to_string()
+    };
+    
+    // EA existence check
+    if !preflight.ea_exists {
+        return Ok(json!({
+            "content": [{ "type": "text", "text": json!({
+                "error": format!("EA '{}' not found in Experts directory.", expert),
+                "hint": "Use search_experts or list_experts to find available EAs."
+            }).to_string() }],
+            "isError": true
+        }));
+    }
+
+    // Date defaulting
+    let (from_date, to_date) = {
+        let f = args.get("from_date").and_then(|v| v.as_str()).unwrap_or("");
+        let t = args.get("to_date").and_then(|v| v.as_str()).unwrap_or("");
+        if f.is_empty() || t.is_empty() {
+            super::past_complete_month()
+        } else {
+            (f.to_string(), t.to_string())
+        }
+    };
+
+    let params = BacktestParams {
+        expert: expert.to_string(),
+        symbol: symbol.to_string(),
+        from_date: from_date.to_string(),
+        to_date: to_date.to_string(),
+        timeframe: args.get("timeframe").and_then(|v| v.as_str()).unwrap_or("M5").to_string(),
+        deposit: args.get("deposit").and_then(|v| v.as_u64()).unwrap_or(10000) as u32,
+        model: args.get("model").and_then(|v| v.as_u64()).unwrap_or(0) as u8,
+        leverage: args.get("leverage").and_then(|v| v.as_u64()).unwrap_or(500) as u32,
+        set_file: args.get("set_file").and_then(|v| v.as_str()).map(|s| s.to_string()),
+        skip_compile: args.get("skip_compile").and_then(|v| v.as_bool()).unwrap_or(false),
+        skip_clean: args.get("skip_clean").and_then(|v| v.as_bool()).unwrap_or(false),
+        skip_analyze: true, // Not needed for launch mode
+        deep_analyze: false,
+        shutdown: false, // Don't shutdown so we can poll
+        kill_existing: false,
+        timeout: args.get("timeout").and_then(|v| v.as_u64()).unwrap_or(900),
+        gui: args.get("gui").and_then(|v| v.as_bool()).unwrap_or(false),
+    };
+
+    let pipeline = BacktestPipeline::new(config.clone());
+    let job = pipeline.launch_backtest(params).await?;
+
+    Ok(json!({
+        "content": [{ "type": "text", "text": json!({
+            "success": true,
+            "message": "Backtest launched successfully. Use get_backtest_status to poll for completion.",
+            "report_id": job.report_id,
+            "report_dir": job.report_dir,
+            "expert": job.expert,
+            "symbol": job.symbol,
+            "timeframe": job.timeframe,
+            "launched_at": job.launched_at,
+            "timeout_seconds": job.timeout_seconds,
+            "poll_hint": "Call get_backtest_status with report_dir to check progress"
+        }).to_string() }],
+        "isError": false
+    }))
+}
+
 pub async fn handle_get_backtest_status(_config: &Config, args: &Value) -> Result<Value> {
    let report_dir = args.get("report_dir")
        .and_then(|v| v.as_str())
        .unwrap_or("latest");

-    let progress_file = Path::new(report_dir).join("progress.log");
+    let report_path = Path::new(report_dir);
+    let progress_file = report_path.join("progress.log");
+    let job_file = report_path.join("job.json");
    
-    let status = if progress_file.exists() {
+    // Load job info if available
+    let job: Option<BacktestJob> = if job_file.exists() {
+        fs::read_to_string(&job_file)
+            .ok()
+            .and_then(|s| serde_json::from_str(&s).ok())
+    } else {
+        None
+    };
+    
+    // Check progress log for stage
+    let (stage, progress_lines) = if progress_file.exists() {
        if let Ok(content) = fs::read_to_string(&progress_file) {
-            let last_line = content.lines().last().unwrap_or("");
-            if last_line.contains("DONE") {
-                "completed"
-            } else {
-                "running"
-            }
+            let lines: Vec<&str> = content.lines().collect();
+            let last_stage = lines.last()
+                .and_then(|l| l.split_whitespace().next())
+                .unwrap_or("UNKNOWN");
+            (last_stage.to_string(), lines.len())
        } else {
-            "unknown"
+            ("UNKNOWN".to_string(), 0)
        }
+    } else {
+        ("NOT_STARTED".to_string(), 0)
+    };
+    
+    // Check if MT5 is running
+    let mt5_running = is_mt5_running();
+    
+    // Check if report file exists
+    let report_found = job.as_ref()
+        .map(|j| Path::new(&j.expected_report_path).exists())
+        .unwrap_or(false);
+    
+    // Check for completed artifacts
+    let metrics_exists = report_path.join("metrics.json").exists();
+    let deals_exists = report_path.join("deals.csv").exists();
+    let is_complete = stage == "DONE" || (report_found && metrics_exists);
+    
+    // Calculate elapsed time if job exists
+    let elapsed_seconds = job.as_ref()
+        .and_then(|j| {
+            chrono::DateTime::parse_from_rfc3339(&j.launched_at)
+                .ok()
+                .map(|t| (chrono::Utc::now() - t.with_timezone(&chrono::Utc)).num_seconds())
+        })
+        .unwrap_or(0);
+    
+    // Determine status message
+    let status_msg = if is_complete {
+        "completed"
+    } else if stage == "BACKTEST" && mt5_running {
+        "running"
+    } else if stage == "BACKTEST" && !mt5_running && !report_found {
+        "failed"
+    } else if progress_lines > 0 {
+        "in_progress"
    } else {
        "not_started"
    };
+    
+    let message = if is_complete {
+        "Backtest completed successfully"
+    } else if stage == "BACKTEST" && mt5_running {
+        "MT5 is running the backtest"
+    } else if stage == "BACKTEST" && !mt5_running {
+        "MT5 process exited but report not found - backtest may have failed"
+    } else {
+        &format!("Backtest is at stage: {}", stage)
+    };

    Ok(json!({
        "content": [{ "type": "text", "text": json!({
            "success": true,
            "report_dir": report_dir,
-            "status": status
+            "status": status_msg,
+            "stage": stage,
+            "is_complete": is_complete,
+            "mt5_running": mt5_running,
+            "report_found": report_found,
+            "metrics_extracted": metrics_exists,
+            "deals_extracted": deals_exists,
+            "elapsed_seconds": elapsed_seconds,
+            "message": message,
+            "job": job.map(|j| {
+                json!({
+                    "report_id": j.report_id,
+                    "expert": j.expert,
+                    "symbol": j.symbol,
+                    "timeframe": j.timeframe,
+                    "launched_at": j.launched_at,
+                    "timeout_seconds": j.timeout_seconds
+                })
+            })
        }).to_string() }],
        "isError": false
    }))
 }

+/// Check if MT5 is currently running
+fn is_mt5_running() -> bool {
+    let patterns = if cfg!(target_os = "macos") {
+        vec!["MetaTrader 5\\.app", "terminal64\\.exe"]
+    } else {
+        vec!["terminal64\\.exe", "metatrader"]
+    };
+    
+    patterns.iter().any(|pat| {
+        Command::new("pgrep")
+            .args(["-f", pat])
+            .output()
+            .map(|o| o.status.success())
+            .unwrap_or(false)
+    })
+}
+
 pub async fn handle_cache_status(config: &Config) -> Result<Value> {
    let cache_dir = config.tester_cache_dir.as_ref()
        .map(|s| Path::new(s))
@@ -4,212 +4,209 @@ use walkdir::WalkDir;
 use crate::compile::MqlCompiler;
 use crate::models::Config;

-pub async fn handle_list_experts(config: &Config, args: &Value) -> Result<Value> {
-    let filter = args.get("filter").and_then(|v| v.as_str());
-    
-    let mut experts = Vec::new();
-    
-    if let Some(experts_dir) = &config.experts_dir {
-        for entry in WalkDir::new(experts_dir).max_depth(3) {
-            if let Ok(entry) = entry {
-                let path = entry.path();
-                if !path.is_file() {
-                    continue;
-                }
-                if let Some(name) = path.file_stem() {
-                    let name_str = name.to_string_lossy().to_string();
-                    let is_compiled = path.extension()
-                        .map(|e| e == "ex5")
-                        .unwrap_or(false);
-                    
-                    if let Some(filter_str) = filter {
-                        if !name_str.to_lowercase().contains(&filter_str.to_lowercase()) {
-                            continue;
-                        }
-                    }
-                    
-                    experts.push(json!({
-                        "name": name_str,
-                        "compiled": is_compiled,
-                        "path": path.to_string_lossy().to_string(),
-                    }));
+const BUILTIN_INDICATORS: &[&str] = &[
+    "Accelerator", "Accumulation", "ADX", "Alligator", "AO", "ATR",
+    "Bands", "Bears", "Bulls", "CCI", "DeMarker", "Envelopes", "Force",
+    "Fractals", "Gator", "Ichimoku", "MA", "MACD", "MFI", "Momentum",
+    "OBV", "OsMA", "RSI", "RVI", "SAR", "StdDev", "Stochastic", "WPR",
+];
+
+// ── Internal helpers ──────────────────────────────────────────────────────────
+
+/// Walk a MQL directory and return file entries matching an optional filter.
+/// `type_label` is included as a `"type"` field when provided (e.g. "custom").
+fn scan_mql_dir(dir: Option<&String>, filter: Option<&str>, type_label: Option<&str>) -> Vec<Value> {
+    let Some(dir) = dir else { return Vec::new() };
+    let filter_lower = filter.map(|f| f.to_lowercase());
+
+    let mut items: Vec<Value> = WalkDir::new(dir)
+        .max_depth(3)
+        .into_iter()
+        .filter_map(|e| e.ok())
+        .filter(|e| e.path().is_file())
+        .filter_map(|e| {
+            let path = e.path();
+            let name = path.file_stem()?.to_string_lossy().into_owned();
+            if let Some(ref f) = filter_lower {
+                if !name.to_lowercase().contains(f.as_str()) {
+                    return None;
                }
            }
-        }
-    }
-    
-    experts.sort_by(|a, b| a["name"].as_str().cmp(&b["name"].as_str()));
-    
-    Ok(json!({
-        "content": [{ "type": "text", "text": json!({
-            "success": true,
-            "count": experts.len(),
-            "experts": experts,
-        }).to_string() }],
+            let is_compiled = path.extension().map(|ext| ext == "ex5").unwrap_or(false);
+            let mut obj = json!({
+                "name": name,
+                "compiled": is_compiled,
+                "path": path.to_string_lossy().as_ref(),
+            });
+            if let Some(t) = type_label {
+                obj["type"] = json!(t);
+            }
+            Some(obj)
+        })
+        .collect();
+
+    items.sort_by(|a, b| a["name"].as_str().cmp(&b["name"].as_str()));
+    items
+}
+
+fn ok_response(data: Value) -> Value {
+    json!({
+        "content": [{ "type": "text", "text": data.to_string() }],
        "isError": false
-    }))
+    })
+}
+
+fn err_response(msg: impl std::fmt::Display) -> Value {
+    json!({
+        "content": [{ "type": "text", "text": msg.to_string() }],
+        "isError": true
+    })
+}
+
+/// Shared implementation for copy_indicator_to_project / copy_script_to_project.
+/// `default_fallback` is used as the stem when the source has no filename.
+fn copy_mql_to_project(config: &Config, args: &Value, default_fallback: &str) -> Result<Value> {
+    use std::path::PathBuf;
+
+    let source_path = args.get("source_path")
+        .and_then(|v| v.as_str())
+        .ok_or_else(|| anyhow::anyhow!("source_path is required"))?;
+
+    let target_name = args.get("target_name").and_then(|v| v.as_str());
+
+    let project_dir = config.project_dir.as_ref()
+        .map(PathBuf::from)
+        .or_else(|| std::env::current_dir().ok())
+        .ok_or_else(|| anyhow::anyhow!("Cannot determine project directory"))?;
+
+    let source = PathBuf::from(source_path);
+    if !source.exists() {
+        return Ok(err_response(
+            serde_json::to_string(&json!({ "success": false, "error": format!("Source file not found: {}", source_path) })).unwrap_or_default()
+        ));
+    }
+
+    let ext = source.extension().and_then(|e| e.to_str()).unwrap_or("mq5");
+    let target_filename = match target_name {
+        Some(name) => format!("{}.{}", name, ext),
+        None => source.file_name()
+            .and_then(|n| n.to_str())
+            .map(|n| n.to_string())
+            .unwrap_or_else(|| format!("{}.{}", default_fallback, ext)),
+    };
+
+    let destination = project_dir.join(&target_filename);
+
+    match std::fs::copy(&source, &destination) {
+        Ok(bytes) => Ok(ok_response(json!({
+            "success": true,
+            "source": source_path,
+            "destination": destination.to_string_lossy().as_ref(),
+            "bytes_copied": bytes,
+        }))),
+        Err(e) => Ok(err_response(
+            serde_json::to_string(&json!({ "success": false, "error": format!("Failed to copy file: {}", e) })).unwrap_or_default()
+        )),
+    }
+}
+
+// ── Public handlers ───────────────────────────────────────────────────────────
+
+pub async fn handle_list_experts(config: &Config, args: &Value) -> Result<Value> {
+    let filter = args.get("filter").and_then(|v| v.as_str());
+    let experts = scan_mql_dir(config.experts_dir.as_ref(), filter, None);
+    Ok(ok_response(json!({
+        "success": true,
+        "count": experts.len(),
+        "experts": experts,
+    })))
 }

 pub async fn handle_search_experts(config: &Config, args: &Value) -> Result<Value> {
    let pattern = args.get("pattern")
        .and_then(|v| v.as_str())
        .ok_or_else(|| anyhow::anyhow!("pattern is required"))?;
-    
-    let pattern_lower = pattern.to_lowercase();
-    let mut matches = Vec::new();
-    
-    if let Some(experts_dir) = &config.experts_dir {
-        for entry in WalkDir::new(experts_dir).max_depth(3) {
-            if let Ok(entry) = entry {
-                let path = entry.path();
-                if !path.is_file() {
-                    continue;
-                }
-                if let Some(name) = path.file_stem() {
-                    let name_str = name.to_string_lossy().to_string();
-                    if name_str.to_lowercase().contains(&pattern_lower) {
-                        let is_compiled = path.extension()
-                            .map(|e| e == "ex5")
-                            .unwrap_or(false);
-                        matches.push(json!({
-                            "name": name_str,
-                            "path": path.to_string_lossy().to_string(),
-                            "compiled": is_compiled,
-                        }));
-                    }
-                }
-            }
-        }
-    }
-    
-    matches.sort_by(|a, b| a["name"].as_str().cmp(&b["name"].as_str()));
-    
-    Ok(json!({
-        "content": [{ "type": "text", "text": json!({
-            "success": true,
-            "pattern": pattern,
-            "count": matches.len(),
-            "matches": matches,
-        }).to_string() }],
-        "isError": false
-    }))
+    let matches = scan_mql_dir(config.experts_dir.as_ref(), Some(pattern), None);
+    Ok(ok_response(json!({
+        "success": true,
+        "pattern": pattern,
+        "count": matches.len(),
+        "matches": matches,
+    })))
 }

 pub async fn handle_list_indicators(config: &Config, args: &Value) -> Result<Value> {
    let filter = args.get("filter").and_then(|v| v.as_str());
    let include_builtin = args.get("include_builtin").and_then(|v| v.as_bool()).unwrap_or(false);
-    
-    let mut indicators = Vec::new();
-    
-    // List custom indicators
-    if let Some(indicators_dir) = &config.indicators_dir {
-        for entry in WalkDir::new(indicators_dir).max_depth(3) {
-            if let Ok(entry) = entry {
-                let path = entry.path();
-                if !path.is_file() {
-                    continue;
-                }
-                if let Some(name) = path.file_stem() {
-                    let name_str = name.to_string_lossy().to_string();
-                    let is_compiled = path.extension()
-                        .map(|e| e == "ex5")
-                        .unwrap_or(false);
-                    
-                    if let Some(filter_str) = filter {
-                        if !name_str.to_lowercase().contains(&filter_str.to_lowercase()) {
-                            continue;
-                        }
-                    }
-                    
-                    indicators.push(json!({
-                        "name": name_str,
-                        "compiled": is_compiled,
-                        "type": "custom",
-                        "path": path.to_string_lossy().to_string(),
-                    }));
-                }
-            }
-        }
-    }
-    
-    // Add built-in indicators if requested
+
+    let mut indicators = scan_mql_dir(config.indicators_dir.as_ref(), filter, Some("custom"));
+
    if include_builtin {
-        let builtin = vec![
-            "Accelerator", "Accumulation", "ADX", "Alligator", "AO", "ATR",
-            "Bands", "Bears", "Bulls", "CCI", "DeMarker", "Envelopes", "Force",
-            "Fractals", "Gator", "Ichimoku", "MA", "MACD", "MFI", "Momentum",
-            "OBV", "OsMA", "RSI", "RVI", "SAR", "StdDev", "Stochastic", "WPR",
-        ];
-        for name in builtin {
-            if filter.map(|f| name.to_lowercase().contains(&f.to_lowercase())).unwrap_or(true) {
-                indicators.push(json!({
-                    "name": name,
-                    "compiled": true,
-                    "type": "builtin",
-                    "path": null,
-                }));
+        let filter_lower = filter.map(|f| f.to_lowercase());
+        for &name in BUILTIN_INDICATORS {
+            if filter_lower.as_ref().map(|f| name.to_lowercase().contains(f.as_str())).unwrap_or(true) {
+                indicators.push(json!({ "name": name, "compiled": true, "type": "builtin", "path": null }));
            }
        }
+        indicators.sort_by(|a, b| a["name"].as_str().cmp(&b["name"].as_str()));
    }
-    
-    indicators.sort_by(|a, b| a["name"].as_str().cmp(&b["name"].as_str()));
-    
-    Ok(json!({
-        "content": [{ "type": "text", "text": json!({
-            "success": true,
-            "count": indicators.len(),
-            "indicators": indicators,
-            "custom_dir": config.indicators_dir.clone(),
-        }).to_string() }],
-        "isError": false
-    }))
+
+    Ok(ok_response(json!({
+        "success": true,
+        "count": indicators.len(),
+        "indicators": indicators,
+        "custom_dir": config.indicators_dir.clone(),
+    })))
 }

 pub async fn handle_list_scripts(config: &Config, args: &Value) -> Result<Value> {
    let filter = args.get("filter").and_then(|v| v.as_str());
-    
-    let mut scripts = Vec::new();
-    
-    if let Some(scripts_dir) = &config.scripts_dir {
-        for entry in WalkDir::new(scripts_dir).max_depth(3) {
-            if let Ok(entry) = entry {
-                let path = entry.path();
-                if !path.is_file() {
-                    continue;
-                }
-                if let Some(name) = path.file_stem() {
-                    let name_str = name.to_string_lossy().to_string();
-                    let is_compiled = path.extension()
-                        .map(|e| e == "ex5")
-                        .unwrap_or(false);
-                    
-                    if let Some(filter_str) = filter {
-                        if !name_str.to_lowercase().contains(&filter_str.to_lowercase()) {
-                            continue;
-                        }
-                    }
-                    
-                    scripts.push(json!({
-                        "name": name_str,
-                        "compiled": is_compiled,
-                        "path": path.to_string_lossy().to_string(),
-                    }));
-                }
+    let scripts = scan_mql_dir(config.scripts_dir.as_ref(), filter, None);
+    Ok(ok_response(json!({
+        "success": true,
+        "count": scripts.len(),
+        "scripts": scripts,
+        "scripts_dir": config.scripts_dir.clone(),
+    })))
+}
+
+pub async fn handle_search_indicators(config: &Config, args: &Value) -> Result<Value> {
+    let pattern = args.get("pattern")
+        .and_then(|v| v.as_str())
+        .ok_or_else(|| anyhow::anyhow!("pattern is required"))?;
+    let include_builtin = args.get("include_builtin").and_then(|v| v.as_bool()).unwrap_or(false);
+
+    let mut matches = scan_mql_dir(config.indicators_dir.as_ref(), Some(pattern), Some("custom"));
+
+    if include_builtin {
+        let pattern_lower = pattern.to_lowercase();
+        for &name in BUILTIN_INDICATORS {
+            if name.to_lowercase().contains(&pattern_lower) {
+                matches.push(json!({ "name": name, "path": null, "type": "builtin", "compiled": true }));
            }
        }
+        matches.sort_by(|a, b| a["name"].as_str().cmp(&b["name"].as_str()));
    }
-    
-    scripts.sort_by(|a, b| a["name"].as_str().cmp(&b["name"].as_str()));
-    
-    Ok(json!({
-        "content": [{ "type": "text", "text": json!({
-            "success": true,
-            "count": scripts.len(),
-            "scripts": scripts,
-            "scripts_dir": config.scripts_dir.clone(),
-        }).to_string() }],
-        "isError": false
-    }))
+
+    Ok(ok_response(json!({
+        "success": true,
+        "pattern": pattern,
+        "count": matches.len(),
+        "matches": matches,
+    })))
+}
+
+pub async fn handle_search_scripts(config: &Config, args: &Value) -> Result<Value> {
+    let pattern = args.get("pattern")
+        .and_then(|v| v.as_str())
+        .ok_or_else(|| anyhow::anyhow!("pattern is required"))?;
+    let matches = scan_mql_dir(config.scripts_dir.as_ref(), Some(pattern), None);
+    Ok(ok_response(json!({
+        "success": true,
+        "pattern": pattern,
+        "count": matches.len(),
+        "matches": matches,
+    })))
 }

 pub async fn handle_compile_ea(config: &Config, args: &Value) -> Result<Value> {
@@ -218,302 +215,49 @@ pub async fn handle_compile_ea(config: &Config, args: &Value) -> Result<Value> {
    let resolved_path: String = if let Some(p) = args.get("expert_path").and_then(|v| v.as_str()) {
        p.to_string()
    } else if let Some(name) = args.get("expert").and_then(|v| v.as_str()) {
-        let mut candidates = vec![
-            PathBuf::from(name).with_extension("mq5"),
-        ];
+        let mut candidates = vec![PathBuf::from(name).with_extension("mq5")];
        if let Some(experts_dir) = &config.experts_dir {
            candidates.push(PathBuf::from(experts_dir).join(name).join(format!("{}.mq5", name)));
            candidates.push(PathBuf::from(experts_dir).join(format!("{}.mq5", name)));
        }
        match candidates.into_iter().find(|p| p.exists()) {
            Some(p) => p.to_string_lossy().to_string(),
-            None => return Ok(serde_json::json!({
-                "content": [{ "type": "text", "text": serde_json::json!({
+            None => return Ok(err_response(
+                serde_json::to_string(&json!({
                    "success": false,
                    "error": format!("Cannot find {}.mq5 in MT5 Experts dir or current directory", name),
-                }).to_string() }],
-                "isError": true
-            })),
+                })).unwrap_or_default()
+            )),
        }
    } else {
        return Err(anyhow::anyhow!("Either 'expert' or 'expert_path' is required"));
    };

    let compiler = MqlCompiler::new(config.clone());
-    let expert_path = resolved_path.as_str();
-    
-    match compiler.compile(&expert_path) {
-        Ok(result) => {
-            Ok(json!({
-                "content": [{ "type": "text", "text": json!({
-                    "success": result.success,
-                    "binary_path": result.ex5_path.map(|p| p.to_string_lossy().to_string()),
-                    "binary_size_bytes": result.binary_size,
-                    "files_synced": result.files_synced,
-                    "warnings": result.warnings.len(),
-                    "errors": result.errors.len(),
-                    "error_list": result.errors,
-                    "warning_list": result.warnings,
-                }).to_string() }],
-                "isError": !result.success
-            }))
-        }
-        Err(e) => {
-            Ok(json!({
-                "content": [{ "type": "text", "text": json!({
-                    "success": false,
-                    "error": format!("Compilation failed: {}", e),
-                }).to_string() }],
-                "isError": true
-            }))
-        }
+    match compiler.compile(&resolved_path).await {
+        Ok(result) => Ok(json!({
+            "content": [{ "type": "text", "text": json!({
+                "success": result.success,
+                "binary_path": result.ex5_path.map(|p| p.to_string_lossy().to_string()),
+                "binary_size_bytes": result.binary_size,
+                "files_synced": result.files_synced,
+                "warnings": result.warnings.len(),
+                "errors": result.errors.len(),
+                "error_list": result.errors,
+                "warning_list": result.warnings,
+            }).to_string() }],
+            "isError": !result.success
+        })),
+        Err(e) => Ok(err_response(
+            serde_json::to_string(&json!({ "success": false, "error": format!("Compilation failed: {}", e) })).unwrap_or_default()
+        )),
    }
 }

-pub async fn handle_search_indicators(config: &Config, args: &Value) -> Result<Value> {
-    let pattern = args.get("pattern")
-        .and_then(|v| v.as_str())
-        .ok_or_else(|| anyhow::anyhow!("pattern is required"))?;
-    
-    let include_builtin = args.get("include_builtin").and_then(|v| v.as_bool()).unwrap_or(false);
-    let pattern_lower = pattern.to_lowercase();
-    
-    let mut matches = Vec::new();
-    
-    // Search custom indicators recursively
-    if let Some(indicators_dir) = &config.indicators_dir {
-        for entry in WalkDir::new(indicators_dir).max_depth(3) {
-            if let Ok(entry) = entry {
-                let path = entry.path();
-                if !path.is_file() {
-                    continue;
-                }
-                if let Some(name) = path.file_stem() {
-                    let name_str = name.to_string_lossy().to_string();
-                    if name_str.to_lowercase().contains(&pattern_lower) {
-                        let is_compiled = path.extension()
-                            .map(|e| e == "ex5")
-                            .unwrap_or(false);
-                        matches.push(json!({
-                            "name": name_str,
-                            "path": path.to_string_lossy().to_string(),
-                            "type": "custom",
-                            "compiled": is_compiled,
-                        }));
-                    }
-                }
-            }
-        }
-    }
-    
-    // Search built-in indicators if requested
-    if include_builtin {
-        let builtin = vec![
-            "Accelerator", "Accumulation", "ADX", "Alligator", "AO", "ATR",
-            "Bands", "Bears", "Bulls", "CCI", "DeMarker", "Envelopes", "Force",
-            "Fractals", "Gator", "Ichimoku", "MA", "MACD", "MFI", "Momentum",
-            "OBV", "OsMA", "RSI", "RVI", "SAR", "StdDev", "Stochastic", "WPR",
-        ];
-        for name in builtin {
-            if name.to_lowercase().contains(&pattern_lower) {
-                matches.push(json!({
-                    "name": name,
-                    "path": null,
-                    "type": "builtin",
-                    "compiled": true,
-                }));
-            }
-        }
-    }
-    
-    matches.sort_by(|a, b| a["name"].as_str().cmp(&b["name"].as_str()));
-    
-    Ok(json!({
-        "content": [{ "type": "text", "text": json!({
-            "success": true,
-            "pattern": pattern,
-            "count": matches.len(),
-            "matches": matches,
-        }).to_string() }],
-        "isError": false
-    }))
-}
-
-pub async fn handle_search_scripts(config: &Config, args: &Value) -> Result<Value> {
-    let pattern = args.get("pattern")
-        .and_then(|v| v.as_str())
-        .ok_or_else(|| anyhow::anyhow!("pattern is required"))?;
-    
-    let pattern_lower = pattern.to_lowercase();
-    let mut matches = Vec::new();
-    
-    if let Some(scripts_dir) = &config.scripts_dir {
-        for entry in WalkDir::new(scripts_dir).max_depth(3) {
-            if let Ok(entry) = entry {
-                let path = entry.path();
-                if !path.is_file() {
-                    continue;
-                }
-                if let Some(name) = path.file_stem() {
-                    let name_str = name.to_string_lossy().to_string();
-                    if name_str.to_lowercase().contains(&pattern_lower) {
-                        let is_compiled = path.extension()
-                            .map(|e| e == "ex5")
-                            .unwrap_or(false);
-                        matches.push(json!({
-                            "name": name_str,
-                            "path": path.to_string_lossy().to_string(),
-                            "compiled": is_compiled,
-                        }));
-                    }
-                }
-            }
-        }
-    }
-    
-    matches.sort_by(|a, b| a["name"].as_str().cmp(&b["name"].as_str()));
-    
-    Ok(json!({
-        "content": [{ "type": "text", "text": json!({
-            "success": true,
-            "pattern": pattern,
-            "count": matches.len(),
-            "matches": matches,
-        }).to_string() }],
-        "isError": false
-    }))
-}
-
 pub async fn handle_copy_indicator_to_project(config: &Config, args: &Value) -> Result<Value> {
-    use std::path::PathBuf;
-    
-    let source_path = args.get("source_path")
-        .and_then(|v| v.as_str())
-        .ok_or_else(|| anyhow::anyhow!("source_path is required"))?;
-    
-    let target_name = args.get("target_name").and_then(|v| v.as_str());
-    
-    // Determine project directory
-    let project_dir = config.project_dir.as_ref()
-        .map(PathBuf::from)
-        .or_else(|| std::env::current_dir().ok())
-        .ok_or_else(|| anyhow::anyhow!("Cannot determine project directory"))?;
-    
-    let source = PathBuf::from(source_path);
-    if !source.exists() {
-        return Ok(json!({
-            "content": [{ "type": "text", "text": json!({
-                "success": false,
-                "error": format!("Source file not found: {}", source_path),
-            }).to_string() }],
-            "isError": true
-        }));
-    }
-    
-    // Get extension
-    let ext = source.extension()
-        .and_then(|e| e.to_str())
-        .unwrap_or("mq5");
-    
-    // Determine target name
-    let target_filename = match target_name {
-        Some(name) => format!("{}.{}", name, ext),
-        None => source.file_name()
-            .and_then(|n| n.to_str())
-            .map(|n| n.to_string())
-            .unwrap_or_else(|| format!("indicator.{}", ext)),
-    };
-    
-    let destination = project_dir.join(&target_filename);
-    
-    // Copy file
-    match std::fs::copy(&source, &destination) {
-        Ok(bytes) => {
-            Ok(json!({
-                "content": [{ "type": "text", "text": json!({
-                    "success": true,
-                    "source": source_path,
-                    "destination": destination.to_string_lossy().to_string(),
-                    "bytes_copied": bytes,
-                }).to_string() }],
-                "isError": false
-            }))
-        }
-        Err(e) => {
-            Ok(json!({
-                "content": [{ "type": "text", "text": json!({
-                    "success": false,
-                    "error": format!("Failed to copy file: {}", e),
-                }).to_string() }],
-                "isError": true
-            }))
-        }
-    }
+    copy_mql_to_project(config, args, "indicator")
 }

 pub async fn handle_copy_script_to_project(config: &Config, args: &Value) -> Result<Value> {
-    use std::path::PathBuf;
-    
-    let source_path = args.get("source_path")
-        .and_then(|v| v.as_str())
-        .ok_or_else(|| anyhow::anyhow!("source_path is required"))?;
-    
-    let target_name = args.get("target_name").and_then(|v| v.as_str());
-    
-    // Determine project directory
-    let project_dir = config.project_dir.as_ref()
-        .map(PathBuf::from)
-        .or_else(|| std::env::current_dir().ok())
-        .ok_or_else(|| anyhow::anyhow!("Cannot determine project directory"))?;
-    
-    let source = PathBuf::from(source_path);
-    if !source.exists() {
-        return Ok(json!({
-            "content": [{ "type": "text", "text": json!({
-                "success": false,
-                "error": format!("Source file not found: {}", source_path),
-            }).to_string() }],
-            "isError": true
-        }));
-    }
-    
-    // Get extension
-    let ext = source.extension()
-        .and_then(|e| e.to_str())
-        .unwrap_or("mq5");
-    
-    // Determine target name
-    let target_filename = match target_name {
-        Some(name) => format!("{}.{}", name, ext),
-        None => source.file_name()
-            .and_then(|n| n.to_str())
-            .map(|n| n.to_string())
-            .unwrap_or_else(|| format!("script.{}", ext)),
-    };
-    
-    let destination = project_dir.join(&target_filename);
-    
-    // Copy file
-    match std::fs::copy(&source, &destination) {
-        Ok(bytes) => {
-            Ok(json!({
-                "content": [{ "type": "text", "text": json!({
-                    "success": true,
-                    "source": source_path,
-                    "destination": destination.to_string_lossy().to_string(),
-                    "bytes_copied": bytes,
-                }).to_string() }],
-                "isError": false
-            }))
-        }
-        Err(e) => {
-            Ok(json!({
-                "content": [{ "type": "text", "text": json!({
-                    "success": false,
-                    "error": format!("Failed to copy file: {}", e),
-                }).to_string() }],
-                "isError": true
-            }))
-        }
-    }
+    copy_mql_to_project(config, args, "script")
 }
@@ -2,6 +2,8 @@ use anyhow::Result;
 use chrono::Datelike;
 use serde_json::{json, Value};
 use std::path::Path;
+use std::sync::OnceLock;
+use std::sync::atomic::{AtomicBool, Ordering};
 use crate::models::Config;

 mod system;
@@ -13,6 +15,13 @@ mod setfiles;
 mod reports;
 mod utility;

+/// Cached result of the background update check.
+/// - Not yet initialized: check still in flight (or not spawned yet)
+/// - Some(version): a newer version is available
+/// - None: already on latest, or GitHub unreachable
+pub(crate) static LATEST_VERSION: OnceLock<Option<String>> = OnceLock::new();
+static BACKGROUND_CHECK_SPAWNED: AtomicBool = AtomicBool::new(false);
+
 #[derive(Debug)]
 pub struct ToolHandler {
    pub config: Config,
@@ -24,12 +33,22 @@ impl ToolHandler {
    }

    pub async fn handle(&self, name: &str, args: &Value) -> Result<Value> {
+        // Spawn a one-shot background update check on the very first tool call of the session.
+        if !BACKGROUND_CHECK_SPAWNED.swap(true, Ordering::Relaxed) {
+            tokio::spawn(async {
+                let result = system::fetch_latest_version().await;
+                let _ = LATEST_VERSION.set(result);
+            });
+        }
+
        match name {
            // System handlers
-            "verify_setup" => system::handle_verify_setup(&self.config).await,
-            "list_symbols" => system::handle_list_symbols(&self.config).await,
-            "healthcheck" => system::handle_healthcheck(&self.config, args).await,
+            "verify_setup"    => system::handle_verify_setup(&self.config).await,
+            "list_symbols"    => system::handle_list_symbols(&self.config).await,
+            "healthcheck"     => system::handle_healthcheck(&self.config, args).await,
            "get_active_account" => system::handle_get_active_account(&self.config).await,
+            "check_update"    => system::handle_check_update(&self.config).await,
+            "update"          => system::handle_update(&self.config).await,
            
            // Expert/Indicator/Script handlers
            "list_experts" => experts::handle_list_experts(&self.config, args).await,
@@ -42,8 +61,11 @@ impl ToolHandler {
            "copy_indicator_to_project" => experts::handle_copy_indicator_to_project(&self.config, args).await,
            "copy_script_to_project" => experts::handle_copy_script_to_project(&self.config, args).await,
            
-            // Backtest handlers
-            "run_backtest" => backtest::handle_run_backtest(&self.config, args).await,
+            // Backtest handlers - Granular pipeline options
+            "run_backtest" => backtest::handle_run_backtest(&self.config, args).await,           // Full: compile + clean + backtest + extract + analyze
+            "run_backtest_quick" => backtest::handle_run_backtest_quick(&self.config, args).await, // Quick: skip compile, do backtest + extract + analyze
+            "run_backtest_only" => backtest::handle_run_backtest_only(&self.config, args).await,  // Minimal: skip compile, do backtest + extract only
+            "launch_backtest" => backtest::handle_launch_backtest(&self.config, args).await,     // Fire-and-forget mode
            "get_backtest_status" => backtest::handle_get_backtest_status(&self.config, args).await,
            "cache_status" => backtest::handle_cache_status(&self.config).await,
            "clean_cache" => backtest::handle_clean_cache(&self.config, args).await,
@@ -65,6 +87,17 @@ impl ToolHandler {
            "analyze_streaks" => analysis::handle_analyze_streaks(&self.config, args).await,
            "analyze_concurrent_peak" => analysis::handle_analyze_concurrent_peak(&self.config, args).await,
            "compare_baseline" => analysis::handle_compare_baseline(&self.config, args).await,
+            // Deal query handlers
+            "list_deals" => analysis::handle_list_deals(&self.config, args).await,
+            "search_deals_by_comment" => analysis::handle_search_deals_by_comment(&self.config, args).await,
+            "search_deals_by_magic" => analysis::handle_search_deals_by_magic(&self.config, args).await,
+            "analyze_profit_distribution" => analysis::handle_analyze_profit_distribution(&self.config, args).await,
+            "analyze_time_performance" => analysis::handle_analyze_time_performance(&self.config, args).await,
+            "analyze_hold_time_distribution" => analysis::handle_analyze_hold_time_distribution(&self.config, args).await,
+            "analyze_layer_performance" => analysis::handle_analyze_layer_performance(&self.config, args).await,
+            "analyze_volume_vs_profit" => analysis::handle_analyze_volume_vs_profit(&self.config, args).await,
+            "analyze_costs" => analysis::handle_analyze_costs(&self.config, args).await,
+            "analyze_efficiency" => analysis::handle_analyze_efficiency(&self.config, args).await,
            
            // Set file handlers
            "read_set_file" => setfiles::handle_read_set_file(args).await,
@@ -87,6 +120,14 @@ impl ToolHandler {
            "promote_to_baseline" => reports::handle_promote_to_baseline(&self.config, args).await,
            "get_history" => reports::handle_get_history(args).await,
            "annotate_history" => reports::handle_annotate_history(args).await,
+            "get_report_by_id" => reports::handle_get_report_by_id(&self.config, args).await,
+            "get_reports_summary" => reports::handle_get_reports_summary(args).await,
+            "get_best_reports" => reports::handle_get_best_reports(args).await,
+            "search_reports_by_tags" => reports::handle_search_reports_by_tags(args).await,
+            "search_reports_by_date_range" => reports::handle_search_reports_by_date_range(args).await,
+            "search_reports_by_notes" => reports::handle_search_reports_by_notes(args).await,
+            "get_reports_by_set_file" => reports::handle_get_reports_by_set_file(args).await,
+            "get_comparable_reports" => reports::handle_get_comparable_reports(args).await,
            
            // Utility handlers
            "check_symbol_data_status" => utility::handle_check_symbol_data_status(&self.config, args).await,
@@ -97,7 +138,17 @@ impl ToolHandler {
            "check_mt5_status" => utility::handle_check_mt5_status(&self.config).await,
            "create_set_template" => utility::handle_create_set_template(&self.config, args).await,
            "export_report" => utility::handle_export_report(&self.config, args).await,
-            
+            // Debugging/diagnostics handlers
+            "diagnose_wine" => utility::handle_diagnose_wine(&self.config, args).await,
+            "get_mt5_logs" => utility::handle_get_mt5_logs(&self.config, args).await,
+            "search_mt5_errors" => utility::handle_search_mt5_errors(&self.config, args).await,
+            "check_mt5_process" => utility::handle_check_mt5_process(&self.config, args).await,
+            "kill_mt5_process" => utility::handle_kill_mt5_process(&self.config, args).await,
+            "check_system_resources" => utility::handle_check_system_resources(&self.config, args).await,
+            "validate_mt5_config" => utility::handle_validate_mt5_config(&self.config, args).await,
+            "get_wine_prefix_info" => utility::handle_get_wine_prefix_info(&self.config, args).await,
+            "get_backtest_crash_info" => utility::handle_get_backtest_crash_info(&self.config, args).await,
+
            _ => Ok(json!({
                "content": [{ "type": "text", "text": format!("Tool '{}' not implemented", name) }],
                "isError": true
@@ -433,3 +433,418 @@ pub async fn handle_annotate_history(args: &Value) -> Result<Value> {
        "isError": false
    }))
 }
+
+pub async fn handle_get_report_by_id(_config: &Config, args: &Value) -> Result<Value> {
+    let id = args.get("id")
+        .and_then(|v| v.as_str())
+        .ok_or_else(|| anyhow::anyhow!("id is required"))?;
+    let include_chart = args.get("include_chart").and_then(|v| v.as_bool()).unwrap_or(true);
+
+    let db = ReportDb::new(&Config::db_path());
+    db.init()?;
+
+    match db.get_by_id(id)? {
+        Some(entry) => {
+            let mut response = json!({
+                "success": true,
+                "report": {
+                    "id": entry.id,
+                    "expert": entry.expert,
+                    "symbol": entry.symbol,
+                    "timeframe": entry.timeframe,
+                    "from_date": entry.from_date,
+                    "to_date": entry.to_date,
+                    "created_at": entry.created_at,
+                    "net_profit": entry.net_profit,
+                    "profit_factor": entry.profit_factor,
+                    "max_dd_pct": entry.max_dd_pct,
+                    "sharpe_ratio": entry.sharpe_ratio,
+                    "total_trades": entry.total_trades,
+                    "win_rate_pct": entry.win_rate_pct,
+                    "recovery_factor": entry.recovery_factor,
+                    "deposit": entry.deposit,
+                    "currency": entry.currency,
+                    "leverage": entry.leverage,
+                    "duration_seconds": entry.duration_seconds,
+                    "set_file_original": entry.set_file_original,
+                    "set_snapshot_path": entry.set_snapshot_path,
+                    "report_dir": entry.report_dir,
+                    "charts_dir": entry.charts_dir,
+                    "tags": entry.tags,
+                    "notes": entry.notes,
+                    "verdict": entry.verdict,
+                }
+            });
+
+            if include_chart {
+                if let Some(charts_dir) = &entry.charts_dir {
+                    let chart_path = Path::new(charts_dir).join("equity.png");
+                    if chart_path.exists() {
+                        match fs::read(&chart_path) {
+                            Ok(bytes) => {
+                                let base64 = BASE64.encode(&bytes);
+                                response["report"]["equity_chart_base64"] = json!(base64);
+                                response["report"]["equity_chart_format"] = json!("png");
+                            }
+                            Err(e) => {
+                                response["report"]["equity_chart_error"] = json!(format!("Failed to read chart: {}", e));
+                            }
+                        }
+                    } else {
+                        response["report"]["equity_chart_error"] = json!("equity.png not found in charts_dir");
+                    }
+                }
+            }
+
+            Ok(json!({
+                "content": [{ "type": "text", "text": response.to_string() }],
+                "isError": false
+            }))
+        }
+        None => {
+            Ok(json!({
+                "content": [{ "type": "text", "text": json!({
+                    "success": false,
+                    "error": format!("Report with id '{}' not found", id)
+                }).to_string() }],
+                "isError": true
+            }))
+        }
+    }
+}
+
+pub async fn handle_get_reports_summary(args: &Value) -> Result<Value> {
+    let db = ReportDb::new(&Config::db_path());
+    db.init()?;
+
+    let filters = ReportFilters {
+        expert: args.get("expert").and_then(|v| v.as_str()).map(|s| s.to_string()),
+        symbol: args.get("symbol").and_then(|v| v.as_str()).map(|s| s.to_string()),
+        timeframe: args.get("timeframe").and_then(|v| v.as_str()).map(|s| s.to_string()),
+        verdict: args.get("verdict").and_then(|v| v.as_str()).map(|s| s.to_string()),
+        ..Default::default()
+    };
+
+    let stats = db.get_stats(&filters)?;
+
+    Ok(json!({
+        "content": [{ "type": "text", "text": json!({
+            "total_count": stats.total_count,
+            "profitable_count": stats.profitable_count,
+            "pass_verdict_count": stats.pass_verdict_count,
+            "fail_verdict_count": stats.fail_verdict_count,
+            "marginal_verdict_count": stats.marginal_verdict_count,
+            "avg_net_profit": stats.avg_net_profit,
+            "avg_profit_factor": stats.avg_profit_factor,
+            "avg_max_dd_pct": stats.avg_max_dd_pct,
+            "avg_win_rate_pct": stats.avg_win_rate_pct,
+            "avg_sharpe_ratio": stats.avg_sharpe_ratio,
+            "profitable_rate": if stats.total_count > 0 {
+                (stats.profitable_count as f64 / stats.total_count as f64 * 100.0).round()
+            } else { 0.0 },
+            "pass_rate": if stats.total_count > 0 {
+                (stats.pass_verdict_count as f64 / stats.total_count as f64 * 100.0).round()
+            } else { 0.0 },
+        }).to_string() }],
+        "isError": false
+    }))
+}
+
+pub async fn handle_get_best_reports(args: &Value) -> Result<Value> {
+    let db = ReportDb::new(&Config::db_path());
+    db.init()?;
+
+    let sort_by = args.get("sort_by").and_then(|v| v.as_str()).unwrap_or("profit_factor");
+    let order = args.get("order").and_then(|v| v.as_str()).unwrap_or("desc");
+    let ascending = order == "asc";
+    let limit = args.get("limit").and_then(|v| v.as_u64()).unwrap_or(10) as usize;
+
+    let filters = ReportFilters {
+        expert: args.get("expert").and_then(|v| v.as_str()).map(|s| s.to_string()),
+        symbol: args.get("symbol").and_then(|v| v.as_str()).map(|s| s.to_string()),
+        timeframe: args.get("timeframe").and_then(|v| v.as_str()).map(|s| s.to_string()),
+        verdict: args.get("verdict").and_then(|v| v.as_str()).map(|s| s.to_string()),
+        ..Default::default()
+    };
+
+    let entries = db.get_sorted_by(sort_by, ascending, limit, &filters)?;
+
+    let reports: Vec<Value> = entries
+        .iter()
+        .map(|e| json!({
+            "id": e.id,
+            "expert": e.expert,
+            "symbol": e.symbol,
+            "timeframe": e.timeframe,
+            "from_date": e.from_date,
+            "to_date": e.to_date,
+            "created_at": e.created_at,
+            "net_profit": e.net_profit,
+            "profit_factor": e.profit_factor,
+            "max_dd_pct": e.max_dd_pct,
+            "sharpe_ratio": e.sharpe_ratio,
+            "total_trades": e.total_trades,
+            "win_rate_pct": e.win_rate_pct,
+            "set_file": e.set_file_original,
+            "verdict": e.verdict,
+            "tags": e.tags,
+        }))
+        .collect();
+
+    Ok(json!({
+        "content": [{ "type": "text", "text": json!({
+            "sort_by": sort_by,
+            "order": order,
+            "matched": reports.len(),
+            "reports": reports,
+        }).to_string() }],
+        "isError": false
+    }))
+}
+
+pub async fn handle_search_reports_by_tags(args: &Value) -> Result<Value> {
+    let tags: Vec<String> = args
+        .get("tags")
+        .and_then(|v| v.as_array())
+        .map(|arr| arr.iter().filter_map(|v| v.as_str().map(|s| s.to_string())).collect())
+        .ok_or_else(|| anyhow::anyhow!("tags array is required"))?;
+
+    let limit = args.get("limit").and_then(|v| v.as_u64()).unwrap_or(50) as usize;
+
+    let db = ReportDb::new(&Config::db_path());
+    db.init()?;
+
+    let entries = db.search_by_tags(&tags, limit)?;
+
+    let reports: Vec<Value> = entries
+        .iter()
+        .map(|e| json!({
+            "id": e.id,
+            "expert": e.expert,
+            "symbol": e.symbol,
+            "timeframe": e.timeframe,
+            "from_date": e.from_date,
+            "to_date": e.to_date,
+            "created_at": e.created_at,
+            "net_profit": e.net_profit,
+            "profit_factor": e.profit_factor,
+            "max_dd_pct": e.max_dd_pct,
+            "total_trades": e.total_trades,
+            "win_rate_pct": e.win_rate_pct,
+            "set_file": e.set_file_original,
+            "verdict": e.verdict,
+            "tags": e.tags,
+            "notes": e.notes,
+        }))
+        .collect();
+
+    Ok(json!({
+        "content": [{ "type": "text", "text": json!({
+            "tags": tags,
+            "matched": reports.len(),
+            "reports": reports,
+        }).to_string() }],
+        "isError": false
+    }))
+}
+
+pub async fn handle_search_reports_by_date_range(args: &Value) -> Result<Value> {
+    let from_start = args.get("from_start").and_then(|v| v.as_str());
+    let from_end = args.get("from_end").and_then(|v| v.as_str());
+    let to_start = args.get("to_start").and_then(|v| v.as_str());
+    let to_end = args.get("to_end").and_then(|v| v.as_str());
+    let limit = args.get("limit").and_then(|v| v.as_u64()).unwrap_or(50) as usize;
+
+    let db = ReportDb::new(&Config::db_path());
+    db.init()?;
+
+    let entries = db.search_by_date_range(from_start, from_end, to_start, to_end, limit)?;
+
+    let reports: Vec<Value> = entries
+        .iter()
+        .map(|e| json!({
+            "id": e.id,
+            "expert": e.expert,
+            "symbol": e.symbol,
+            "timeframe": e.timeframe,
+            "from_date": e.from_date,
+            "to_date": e.to_date,
+            "created_at": e.created_at,
+            "net_profit": e.net_profit,
+            "profit_factor": e.profit_factor,
+            "max_dd_pct": e.max_dd_pct,
+            "total_trades": e.total_trades,
+            "win_rate_pct": e.win_rate_pct,
+            "set_file": e.set_file_original,
+            "verdict": e.verdict,
+            "tags": e.tags,
+        }))
+        .collect();
+
+    Ok(json!({
+        "content": [{ "type": "text", "text": json!({
+            "from_start": from_start,
+            "from_end": from_end,
+            "to_start": to_start,
+            "to_end": to_end,
+            "matched": reports.len(),
+            "reports": reports,
+        }).to_string() }],
+        "isError": false
+    }))
+}
+
+pub async fn handle_search_reports_by_notes(args: &Value) -> Result<Value> {
+    let query = args.get("query")
+        .and_then(|v| v.as_str())
+        .ok_or_else(|| anyhow::anyhow!("query is required"))?;
+    let limit = args.get("limit").and_then(|v| v.as_u64()).unwrap_or(50) as usize;
+
+    let db = ReportDb::new(&Config::db_path());
+    db.init()?;
+
+    let entries = db.search_by_notes(query, limit)?;
+
+    let reports: Vec<Value> = entries
+        .iter()
+        .map(|e| json!({
+            "id": e.id,
+            "expert": e.expert,
+            "symbol": e.symbol,
+            "timeframe": e.timeframe,
+            "from_date": e.from_date,
+            "to_date": e.to_date,
+            "created_at": e.created_at,
+            "net_profit": e.net_profit,
+            "profit_factor": e.profit_factor,
+            "max_dd_pct": e.max_dd_pct,
+            "total_trades": e.total_trades,
+            "win_rate_pct": e.win_rate_pct,
+            "set_file": e.set_file_original,
+            "verdict": e.verdict,
+            "notes": e.notes,
+            "tags": e.tags,
+        }))
+        .collect();
+
+    Ok(json!({
+        "content": [{ "type": "text", "text": json!({
+            "query": query,
+            "matched": reports.len(),
+            "reports": reports,
+        }).to_string() }],
+        "isError": false
+    }))
+}
+
+pub async fn handle_get_reports_by_set_file(args: &Value) -> Result<Value> {
+    let set_file = args.get("set_file")
+        .and_then(|v| v.as_str())
+        .ok_or_else(|| anyhow::anyhow!("set_file is required"))?;
+    let limit = args.get("limit").and_then(|v| v.as_u64()).unwrap_or(50) as usize;
+
+    let db = ReportDb::new(&Config::db_path());
+    db.init()?;
+
+    let entries = db.search_by_set_file(set_file, limit)?;
+
+    let reports: Vec<Value> = entries
+        .iter()
+        .map(|e| json!({
+            "id": e.id,
+            "expert": e.expert,
+            "symbol": e.symbol,
+            "timeframe": e.timeframe,
+            "from_date": e.from_date,
+            "to_date": e.to_date,
+            "created_at": e.created_at,
+            "net_profit": e.net_profit,
+            "profit_factor": e.profit_factor,
+            "max_dd_pct": e.max_dd_pct,
+            "total_trades": e.total_trades,
+            "win_rate_pct": e.win_rate_pct,
+            "set_file_original": e.set_file_original,
+            "set_snapshot_path": e.set_snapshot_path,
+            "verdict": e.verdict,
+            "tags": e.tags,
+        }))
+        .collect();
+
+    Ok(json!({
+        "content": [{ "type": "text", "text": json!({
+            "set_file": set_file,
+            "matched": reports.len(),
+            "reports": reports,
+        }).to_string() }],
+        "isError": false
+    }))
+}
+
+pub async fn handle_get_comparable_reports(args: &Value) -> Result<Value> {
+    let db = ReportDb::new(&Config::db_path());
+    db.init()?;
+
+    // Get expert/symbol/timeframe either from report_id or direct args
+    let (expert, symbol, timeframe, exclude_id) = if let Some(id) = args.get("report_id").and_then(|v| v.as_str()) {
+        match db.get_by_id(id)? {
+            Some(entry) => {
+                let exclude = args.get("exclude_id").and_then(|v| v.as_str()).map(|s| s.to_string());
+                (entry.expert, entry.symbol, entry.timeframe, exclude.unwrap_or_else(|| id.to_string()))
+            }
+            None => return Ok(json!({
+                "content": [{ "type": "text", "text": json!({
+                    "success": false,
+                    "error": format!("Report with id '{}' not found", id)
+                }).to_string() }],
+                "isError": true
+            }))
+        }
+    } else {
+        let expert = args.get("expert").and_then(|v| v.as_str())
+            .ok_or_else(|| anyhow::anyhow!("expert or report_id is required"))?;
+        let symbol = args.get("symbol").and_then(|v| v.as_str())
+            .ok_or_else(|| anyhow::anyhow!("symbol or report_id is required"))?;
+        let timeframe = args.get("timeframe").and_then(|v| v.as_str())
+            .ok_or_else(|| anyhow::anyhow!("timeframe or report_id is required"))?;
+        let exclude_id = args.get("exclude_id").and_then(|v| v.as_str()).map(|s| s.to_string());
+        (expert.to_string(), symbol.to_string(), timeframe.to_string(), exclude_id.unwrap_or_default())
+    };
+
+    let limit = args.get("limit").and_then(|v| v.as_u64()).unwrap_or(20) as usize;
+    let exclude_opt = if exclude_id.is_empty() { None } else { Some(exclude_id.as_str()) };
+
+    let entries = db.get_comparable(&expert, &symbol, &timeframe, exclude_opt, limit)?;
+
+    let reports: Vec<Value> = entries
+        .iter()
+        .map(|e| json!({
+            "id": e.id,
+            "expert": e.expert,
+            "symbol": e.symbol,
+            "timeframe": e.timeframe,
+            "from_date": e.from_date,
+            "to_date": e.to_date,
+            "created_at": e.created_at,
+            "net_profit": e.net_profit,
+            "profit_factor": e.profit_factor,
+            "max_dd_pct": e.max_dd_pct,
+            "total_trades": e.total_trades,
+            "win_rate_pct": e.win_rate_pct,
+            "set_file": e.set_file_original,
+            "verdict": e.verdict,
+            "tags": e.tags,
+        }))
+        .collect();
+
+    Ok(json!({
+        "content": [{ "type": "text", "text": json!({
+            "expert": expert,
+            "symbol": symbol,
+            "timeframe": timeframe,
+            "exclude_id": exclude_id,
+            "matched": reports.len(),
+            "reports": reports,
+        }).to_string() }],
+        "isError": false
+    }))
+}
@@ -3,6 +3,178 @@ use serde_json::{json, Value};
 use std::path::Path;
 use crate::models::Config;

+// ── Update helpers ────────────────────────────────────────────────────────────
+
+fn platform_tag() -> &'static str {
+    #[cfg(all(target_os = "macos", target_arch = "aarch64"))] return "macos-aarch64";
+    #[cfg(all(target_os = "macos", target_arch = "x86_64"))]  return "macos-x86_64";
+    #[cfg(all(target_os = "linux", target_arch = "x86_64"))]  return "linux-x86_64";
+    #[cfg(not(any(
+        all(target_os = "macos", target_arch = "aarch64"),
+        all(target_os = "macos", target_arch = "x86_64"),
+        all(target_os = "linux", target_arch = "x86_64"),
+    )))] return "unsupported";
+}
+
+fn semver_newer(latest: &str, current: &str) -> bool {
+    let parse = |s: &str| -> (u32, u32, u32) {
+        let mut p = s.trim_start_matches('v').splitn(3, '.');
+        let ma = p.next().and_then(|x| x.parse().ok()).unwrap_or(0);
+        let mi = p.next().and_then(|x| x.parse().ok()).unwrap_or(0);
+        let pa = p.next().and_then(|x| x.parse().ok()).unwrap_or(0);
+        (ma, mi, pa)
+    };
+    parse(latest) > parse(current)
+}
+
+/// Fetch the latest release tag from GitHub API (5 s timeout via curl).
+/// Returns the version string without the leading "v", or None on failure.
+pub(super) async fn fetch_latest_version() -> Option<String> {
+    let output = tokio::process::Command::new("curl")
+        .args([
+            "-sf", "--max-time", "5",
+            "-H", "Accept: application/vnd.github.v3+json",
+            "-H", "User-Agent: mt5-quant-updater",
+            "https://api.github.com/repos/masdevid/mt5-quant/releases/latest",
+        ])
+        .output()
+        .await
+        .ok()?;
+
+    if !output.status.success() { return None; }
+
+    let body: Value = serde_json::from_slice(&output.stdout).ok()?;
+    body.get("tag_name")
+        .and_then(|v| v.as_str())
+        .map(|s| s.trim_start_matches('v').to_string())
+}
+
+fn ok_response(data: Value) -> Value {
+    json!({ "content": [{ "type": "text", "text": data.to_string() }], "isError": false })
+}
+
+fn err_response(msg: impl std::fmt::Display) -> Value {
+    json!({ "content": [{ "type": "text", "text": msg.to_string() }], "isError": true })
+}
+
+// ── Update tool handlers ──────────────────────────────────────────────────────
+
+pub async fn handle_check_update(_config: &Config) -> Result<Value> {
+    let current = env!("CARGO_PKG_VERSION");
+
+    // Use cached background-check result if available; otherwise fetch now.
+    let latest_opt = match super::LATEST_VERSION.get() {
+        Some(v) => v.clone(),
+        None => fetch_latest_version().await,
+    };
+
+    let Some(latest) = latest_opt else {
+        return Ok(ok_response(json!({
+            "current_version": current,
+            "update_available": false,
+            "error": "Could not reach GitHub API — check network connectivity",
+        })));
+    };
+
+    let update_available = semver_newer(&latest, current);
+    Ok(ok_response(json!({
+        "current_version": current,
+        "latest_version": latest,
+        "update_available": update_available,
+        "hint": if update_available {
+            format!("Run the `update` tool to install v{latest}")
+        } else {
+            "You are on the latest version".to_string()
+        },
+    })))
+}
+
+pub async fn handle_update(_config: &Config) -> Result<Value> {
+    let current = env!("CARGO_PKG_VERSION");
+
+    let latest = match super::LATEST_VERSION.get().and_then(|v| v.as_deref()) {
+        Some(v) => v.to_string(),
+        None => match fetch_latest_version().await {
+            Some(v) => v,
+            None => return Ok(err_response(
+                r#"{"success":false,"error":"Could not determine latest version — check network"}"#
+            )),
+        },
+    };
+
+    if !semver_newer(&latest, current) {
+        return Ok(ok_response(json!({
+            "up_to_date": true,
+            "version": current,
+        })));
+    }
+
+    let tag = platform_tag();
+    if tag == "unsupported" {
+        return Ok(err_response(
+            r#"{"success":false,"error":"Auto-update not supported on this platform — build from source"}"#
+        ));
+    }
+
+    let url = format!(
+        "https://github.com/masdevid/mt5-quant/releases/download/v{latest}/mcp-mt5-quant-{tag}.tar.gz"
+    );
+
+    // Download tarball to a temp file
+    let tmp_tar = tempfile::NamedTempFile::new()?;
+    let dl = tokio::process::Command::new("curl")
+        .args(["-sfL", "--max-time", "120",
+               "-o", tmp_tar.path().to_str().unwrap_or_default(),
+               &url])
+        .status()
+        .await?;
+
+    if !dl.success() {
+        return Ok(err_response(format!(
+            r#"{{"success":false,"error":"Download failed","url":"{}"}}"#, url
+        )));
+    }
+
+    // Extract binary (tarball root dir is mcp-mt5-quant-{platform}/)
+    let tmp_dir = tempfile::tempdir()?;
+    let extract = tokio::process::Command::new("tar")
+        .args(["-xzf", tmp_tar.path().to_str().unwrap_or_default(),
+               "-C", tmp_dir.path().to_str().unwrap_or_default(),
+               "--strip-components=1"])
+        .status()
+        .await?;
+
+    if !extract.success() {
+        return Ok(err_response(r#"{"success":false,"error":"Failed to extract archive"}"#));
+    }
+
+    let new_bin = tmp_dir.path().join("mt5-quant");
+    if !new_bin.exists() {
+        return Ok(err_response(r#"{"success":false,"error":"Binary not found in archive"}"#));
+    }
+
+    // Atomic replace: write to sibling .tmp, then rename (safe on same FS)
+    let current_exe = std::env::current_exe()?;
+    let tmp_dest = current_exe.with_extension("update_tmp");
+    std::fs::copy(&new_bin, &tmp_dest)?;
+
+    #[cfg(unix)]
+    {
+        use std::os::unix::fs::PermissionsExt;
+        std::fs::set_permissions(&tmp_dest, std::fs::Permissions::from_mode(0o755))?;
+    }
+
+    std::fs::rename(&tmp_dest, &current_exe)?;
+
+    Ok(ok_response(json!({
+        "success": true,
+        "previous_version": current,
+        "updated_to": latest,
+        "binary": current_exe.to_string_lossy(),
+        "hint": format!("Updated to v{latest}. Restart the MCP connection to load the new binary."),
+    })))
+}
+
 pub async fn handle_verify_setup(config: &Config) -> Result<Value> {
    let mut checks = serde_json::Map::new();
    let mut all_ok = true;
@@ -841,3 +841,878 @@ pub async fn handle_export_report(_config: &Config, args: &Value) -> Result<Valu
        "isError": false
    }))
 }
+
+// === Wine & MT5 Debugging Tools ===
+
+/// Diagnose Wine installation and prefix health
+pub async fn handle_diagnose_wine(config: &Config, _args: &Value) -> Result<Value> {
+    let mut diagnostics = json!({
+        "wine_executable": null,
+        "wine_version": null,
+        "wine_prefix": null,
+        "prefix_health": null,
+        "prefix_exists": false,
+        "prefix_size_mb": 0,
+        "errors": Vec::<String>::new(),
+        "warnings": Vec::<String>::new(),
+    });
+    
+    // Check wine executable
+    if let Some(wine_exe) = config.wine_executable.as_ref() {
+        diagnostics["wine_executable"] = json!(wine_exe);
+        
+        // Get Wine version
+        let version_output = std::process::Command::new(wine_exe)
+            .arg("--version")
+            .output();
+        
+        match version_output {
+            Ok(output) if output.status.success() => {
+                let version = String::from_utf8_lossy(&output.stdout).trim().to_string();
+                diagnostics["wine_version"] = json!(version);
+            }
+            _ => {
+                diagnostics["errors"].as_array_mut().unwrap().push(
+                    json!("Failed to get Wine version - Wine may not be properly installed")
+                );
+            }
+        }
+    } else {
+        diagnostics["errors"].as_array_mut().unwrap().push(
+            json!("Wine executable not configured")
+        );
+    }
+    
+    // Check Wine prefix
+    if let Some(mt5_dir) = config.mt5_dir() {
+        let wine_prefix = mt5_dir
+            .parent()
+            .and_then(|p| p.parent())
+            .and_then(|p| p.parent());
+        
+        if let Some(prefix) = wine_prefix {
+            diagnostics["wine_prefix"] = json!(prefix.to_string_lossy().to_string());
+            
+            // Check if prefix exists
+            let prefix_exists = prefix.exists();
+            diagnostics["prefix_exists"] = json!(prefix_exists);
+            
+            if prefix_exists {
+                // Calculate prefix size
+                let mut total_size = 0u64;
+                fn calculate_size(dir: &Path, total: &mut u64) {
+                    if let Ok(entries) = fs::read_dir(dir) {
+                        for entry in entries.flatten() {
+                            let path = entry.path();
+                            if path.is_file() {
+                                if let Ok(meta) = entry.metadata() {
+                                    *total += meta.len();
+                                }
+                            } else if path.is_dir() {
+                                calculate_size(&path, total);
+                            }
+                        }
+                    }
+                }
+                calculate_size(prefix, &mut total_size);
+                diagnostics["prefix_size_mb"] = json!((total_size / 1024 / 1024) as i64);
+                
+                // Check critical directories
+                let system32 = prefix.join("drive_c/windows/system32");
+                let program_files = prefix.join("drive_c/Program Files");
+                
+                if !system32.exists() {
+                    diagnostics["errors"].as_array_mut().unwrap().push(
+                        json!("Wine prefix missing system32 directory - prefix may be corrupted")
+                    );
+                    diagnostics["prefix_health"] = json!("corrupted");
+                } else if !program_files.exists() {
+                    diagnostics["warnings"].as_array_mut().unwrap().push(
+                        json!("Program Files directory not found")
+                    );
+                    diagnostics["prefix_health"] = json!("incomplete");
+                } else {
+                    diagnostics["prefix_health"] = json!("healthy");
+                }
+                
+                // Check for recent Wine errors
+                let wine_log = prefix.join("wine.log");
+                if wine_log.exists() {
+                    if let Ok(content) = fs::read_to_string(&wine_log) {
+                        let recent_errors: Vec<&str> = content.lines()
+                            .filter(|l| l.contains("err:") || l.contains("fixme:"))
+                            .rev()
+                            .take(10)
+                            .collect();
+                        if !recent_errors.is_empty() {
+                            diagnostics["recent_wine_errors"] = json!(recent_errors);
+                        }
+                    }
+                }
+            } else {
+                diagnostics["errors"].as_array_mut().unwrap().push(
+                    json!("Wine prefix directory does not exist")
+                );
+                diagnostics["prefix_health"] = json!("missing");
+            }
+        } else {
+            diagnostics["errors"].as_array_mut().unwrap().push(
+                json!("Could not determine Wine prefix from MT5 directory")
+            );
+        }
+    } else {
+        diagnostics["errors"].as_array_mut().unwrap().push(
+            json!("MT5 directory not configured")
+        );
+    }
+    
+    let has_errors = !diagnostics["errors"].as_array().unwrap().is_empty();
+    
+    Ok(json!({
+        "content": [{ "type": "text", "text": diagnostics.to_string() }],
+        "isError": has_errors
+    }))
+}
+
+/// Get MT5 terminal logs
+pub async fn handle_get_mt5_logs(config: &Config, args: &Value) -> Result<Value> {
+    let log_type = args.get("log_type").and_then(|v| v.as_str()).unwrap_or("terminal");
+    let lines = args.get("lines").and_then(|v| v.as_u64()).unwrap_or(100) as usize;
+    let search = args.get("search").and_then(|v| v.as_str());
+    
+    let mt5_dir = config.mt5_dir()
+        .ok_or_else(|| anyhow::anyhow!("MT5 directory not configured"))?;
+    
+    let log_path = match log_type {
+        "terminal" => mt5_dir.join("logs").join(format!("{}", chrono::Local::now().format("%Y%m%d"))),
+        "tester" => mt5_dir.join("Tester").join("logs"),
+        "metaeditor" => mt5_dir.join("MetaEditor").join("logs"),
+        _ => mt5_dir.join("logs"),
+    };
+    
+    let mut result = json!({
+        "log_type": log_type,
+        "log_path": log_path.to_string_lossy().to_string(),
+        "found": false,
+        "lines_total": 0,
+        "lines_returned": 0,
+        "content": Vec::<String>::new(),
+    });
+    
+    // Find log files
+    let mut log_files: Vec<_> = Vec::new();
+    if log_path.exists() {
+        if let Ok(entries) = fs::read_dir(&log_path) {
+            for entry in entries.flatten() {
+                let path = entry.path();
+                if path.is_file() {
+                    if let Some(ext) = path.extension() {
+                        if ext == "log" {
+                            if let Ok(meta) = entry.metadata() {
+                                if let Ok(modified) = meta.modified() {
+                                    log_files.push((path, modified));
+                                }
+                            }
+                        }
+                    }
+                }
+            }
+        }
+    }
+    
+    // Sort by modification time (newest first)
+    log_files.sort_by(|a, b| b.1.cmp(&a.1));
+    
+    if let Some((latest_log, _)) = log_files.first() {
+        result["found"] = json!(true);
+        
+        if let Ok(content) = fs::read_to_string(latest_log) {
+            let all_lines: Vec<&str> = content.lines().collect();
+            result["lines_total"] = json!(all_lines.len());
+            
+            // Filter and limit lines
+            let mut filtered: Vec<&str> = all_lines.clone();
+            
+            // Apply search filter
+            if let Some(search_term) = search {
+                let search_lower = search_term.to_lowercase();
+                filtered.retain(|line| line.to_lowercase().contains(&search_lower));
+            }
+            
+            // Get last N lines
+            let start = filtered.len().saturating_sub(lines);
+            let final_lines: Vec<String> = filtered[start..].iter().map(|s| s.to_string()).collect();
+            
+            result["lines_returned"] = json!(final_lines.len());
+            result["content"] = json!(final_lines);
+        }
+    }
+    
+    Ok(json!({
+        "content": [{ "type": "text", "text": result.to_string() }],
+        "isError": false
+    }))
+}
+
+/// Search MT5 logs for error patterns
+pub async fn handle_search_mt5_errors(config: &Config, args: &Value) -> Result<Value> {
+    let error_patterns = vec![
+        "error", "failed", "crash", "exception", "access violation",
+        "out of memory", "cannot", "unable to", "terminated"
+    ];
+    let hours_back = args.get("hours_back").and_then(|v| v.as_u64()).unwrap_or(24);
+    let max_errors = args.get("max_errors").and_then(|v| v.as_u64()).unwrap_or(50) as usize;
+    
+    let mt5_dir = config.mt5_dir()
+        .ok_or_else(|| anyhow::anyhow!("MT5 directory not configured"))?;
+    
+    let mut errors_found = Vec::new();
+    let logs_dir = mt5_dir.join("logs");
+    let cutoff_time = std::time::SystemTime::now() - std::time::Duration::from_secs(hours_back * 3600);
+    
+    // Search recent log files
+    if logs_dir.exists() {
+        if let Ok(entries) = fs::read_dir(&logs_dir) {
+            for entry in entries.flatten() {
+                let path = entry.path();
+                if path.extension().map(|e| e == "log").unwrap_or(false) {
+                    if let Ok(meta) = entry.metadata() {
+                        if let Ok(modified) = meta.modified() {
+                            if modified >= cutoff_time {
+                                if let Ok(content) = fs::read_to_string(&path) {
+                                    for (i, line) in content.lines().enumerate() {
+                                        let line_lower = line.to_lowercase();
+                                        for pattern in &error_patterns {
+                                            if line_lower.contains(pattern) {
+                                                errors_found.push(json!({
+                                                    "file": path.file_name().unwrap_or_default().to_string_lossy().to_string(),
+                                                    "line": i + 1,
+                                                    "content": line.trim().to_string(),
+                                                    "pattern": pattern,
+                                                }));
+                                                if errors_found.len() >= max_errors {
+                                                    break;
+                                                }
+                                            }
+                                        }
+                                    }
+                                }
+                            }
+                        }
+                    }
+                }
+                if errors_found.len() >= max_errors {
+                    break;
+                }
+            }
+        }
+    }
+    
+    let result = json!({
+        "hours_searched": hours_back,
+        "errors_found": errors_found.len(),
+        "max_errors": max_errors,
+        "errors": errors_found,
+        "suggestion": if errors_found.is_empty() {
+            "No errors found in recent logs. Check get_mt5_logs for full log content."
+        } else {
+            "Found potential errors. Review the 'content' field for details."
+        },
+    });
+    
+    Ok(json!({
+        "content": [{ "type": "text", "text": result.to_string() }],
+        "isError": false
+    }))
+}
+
+/// Check MT5 process status
+pub async fn handle_check_mt5_process(_config: &Config, _args: &Value) -> Result<Value> {
+    use std::process::Command;
+    
+    let mut result = json!({
+        "is_running": false,
+        "processes": Vec::<serde_json::Value>::new(),
+        "wine_server_running": false,
+        "total_instances": 0,
+    });
+    
+    #[cfg(target_os = "macos")]
+    {
+        // Check for MT5 processes
+        let ps_output = Command::new("ps")
+            .args(["aux"])
+            .output();
+        
+        if let Ok(output) = ps_output {
+            let content = String::from_utf8_lossy(&output.stdout);
+            let mut processes = Vec::new();
+            let mut mt5_count = 0;
+            let mut wine_server = false;
+            
+            for line in content.lines() {
+                let line_lower = line.to_lowercase();
+                
+                if line_lower.contains("terminal64") || line_lower.contains("metatrader") {
+                    let parts: Vec<&str> = line.split_whitespace().collect();
+                    if parts.len() >= 11 {
+                        processes.push(json!({
+                            "pid": parts[1],
+                            "cpu": parts[2],
+                            "mem": parts[3],
+                            "command": parts[10..].join(" "),
+                        }));
+                        mt5_count += 1;
+                    }
+                }
+                
+                if line_lower.contains("wineserver") {
+                    wine_server = true;
+                }
+            }
+            
+            result["processes"] = json!(processes);
+            result["is_running"] = json!(mt5_count > 0);
+            result["total_instances"] = json!(mt5_count);
+            result["wine_server_running"] = json!(wine_server);
+        }
+    }
+    
+    #[cfg(target_os = "linux")]
+    {
+        let ps_output = Command::new("ps")
+            .args(["aux"])
+            .output();
+        
+        if let Ok(output) = ps_output {
+            let content = String::from_utf8_lossy(&output.stdout);
+            let mut processes = Vec::new();
+            let mut mt5_count = 0;
+            let mut wine_server = false;
+            
+            for line in content.lines() {
+                let line_lower = line.to_lowercase();
+                
+                if line_lower.contains("terminal64") || line_lower.contains("metatrader") {
+                    let parts: Vec<&str> = line.split_whitespace().collect();
+                    if parts.len() >= 11 {
+                        processes.push(json!({
+                            "pid": parts[1],
+                            "cpu": parts[2],
+                            "mem": parts[3],
+                            "command": parts[10..].join(" "),
+                        }));
+                        mt5_count += 1;
+                    }
+                }
+                
+                if line_lower.contains("wineserver") {
+                    wine_server = true;
+                }
+            }
+            
+            result["processes"] = json!(processes);
+            result["is_running"] = json!(mt5_count > 0);
+            result["total_instances"] = json!(mt5_count);
+            result["wine_server_running"] = json!(wine_server);
+        }
+    }
+    
+    Ok(json!({
+        "content": [{ "type": "text", "text": result.to_string() }],
+        "isError": false
+    }))
+}
+
+/// Kill stuck MT5 process
+pub async fn handle_kill_mt5_process(_config: &Config, args: &Value) -> Result<Value> {
+    use std::process::Command;
+    
+    let force = args.get("force").and_then(|v| v.as_bool()).unwrap_or(false);
+    let pid = args.get("pid").and_then(|v| v.as_str());
+    
+    let mut killed = Vec::new();
+    let mut failed = Vec::new();
+    
+    #[cfg(any(target_os = "macos", target_os = "linux"))]
+    {
+        // Get list of MT5 processes
+        let ps_output = Command::new("ps")
+            .args(["aux"])
+            .output();
+        
+        if let Ok(output) = ps_output {
+            let content = String::from_utf8_lossy(&output.stdout);
+            
+            for line in content.lines() {
+                let line_lower = line.to_lowercase();
+                
+                let should_kill = if let Some(target_pid) = pid {
+                    line.contains(target_pid) && (line_lower.contains("terminal64") || line_lower.contains("metatrader"))
+                } else {
+                    line_lower.contains("terminal64") || line_lower.contains("metatrader")
+                };
+                
+                if should_kill {
+                    let parts: Vec<&str> = line.split_whitespace().collect();
+                    if parts.len() >= 2 {
+                        let process_pid = parts[1];
+                        let signal = if force { "-9" } else { "-15" };
+                        
+                        match Command::new("kill").args([signal, process_pid]).output() {
+                            Ok(_) => killed.push(process_pid.to_string()),
+                            Err(e) => failed.push(format!("{}: {}", process_pid, e)),
+                        }
+                    }
+                }
+            }
+        }
+        
+        // Also kill wineserver if force=true
+        if force {
+            let _ = Command::new("killall").arg("wineserver").output();
+        }
+    }
+    
+    let message = if killed.is_empty() {
+        "No MT5 processes found to kill".to_string()
+    } else {
+        format!("Killed {} MT5 process(es)", killed.len())
+    };
+    
+    let result = json!({
+        "killed": killed,
+        "failed": failed,
+        "force": force,
+        "message": message,
+    });
+    
+    Ok(json!({
+        "content": [{ "type": "text", "text": result.to_string() }],
+        "isError": !failed.is_empty()
+    }))
+}
+
+/// Check system resources for MT5
+pub async fn handle_check_system_resources(_config: &Config, _args: &Value) -> Result<Value> {
+    use std::process::Command;
+    
+    let mut result = json!({
+        "disk_space": null,
+        "memory": null,
+        "cpu_cores": 0,
+        "recommendations": Vec::<String>::new(),
+    });
+    
+    #[cfg(any(target_os = "macos", target_os = "linux"))]
+    {
+        // Check disk space
+        let df_output = Command::new("df")
+            .args(["-h", "/"])
+            .output();
+        
+        if let Ok(output) = df_output {
+            let content = String::from_utf8_lossy(&output.stdout);
+            for line in content.lines().skip(1) {
+                let parts: Vec<&str> = line.split_whitespace().collect();
+                if parts.len() >= 5 {
+                    result["disk_space"] = json!({
+                        "filesystem": parts[0],
+                        "size": parts[1],
+                        "used": parts[2],
+                        "available": parts[3],
+                        "use_percent": parts[4],
+                    });
+                    
+                    // Check if low on space
+                    let use_pct = parts[4].trim_end_matches('%').parse::<u32>().unwrap_or(0);
+                    if use_pct > 90 {
+                        result["recommendations"].as_array_mut().unwrap().push(
+                            json!("Disk space critically low. Clean MT5 cache with clean_cache.")
+                        );
+                    } else if use_pct > 80 {
+                        result["recommendations"].as_array_mut().unwrap().push(
+                            json!("Disk space getting low. Consider cleaning cache.")
+                        );
+                    }
+                }
+            }
+        }
+        
+        // Check memory
+        #[cfg(target_os = "macos")]
+        {
+            let vm_output = Command::new("vm_stat").output();
+            if let Ok(output) = vm_output {
+                let content = String::from_utf8_lossy(&output.stdout);
+                // Parse vm_stat output
+                let mut free_pages = 0u64;
+                let mut active_pages = 0u64;
+                let mut inactive_pages = 0u64;
+                
+                for line in content.lines() {
+                    if line.contains("Pages free:") {
+                        free_pages = line.split_whitespace().nth(2).unwrap_or("0").trim_end_matches('.').parse().unwrap_or(0);
+                    } else if line.contains("Pages active:") {
+                        active_pages = line.split_whitespace().nth(2).unwrap_or("0").trim_end_matches('.').parse().unwrap_or(0);
+                    } else if line.contains("Pages inactive:") {
+                        inactive_pages = line.split_whitespace().nth(2).unwrap_or("0").trim_end_matches('.').parse().unwrap_or(0);
+                    }
+                }
+                
+                let page_size = 4096u64;
+                let total_mb = ((free_pages + active_pages + inactive_pages) * page_size) / 1024 / 1024;
+                let free_mb = (free_pages * page_size) / 1024 / 1024;
+                
+                result["memory"] = json!({
+                    "total_mb": total_mb,
+                    "free_mb": free_mb,
+                    "unit": "MB",
+                });
+                
+                if free_mb < 2048 {
+                    result["recommendations"].as_array_mut().unwrap().push(
+                        json!("Low memory available. MT5 may crash during large optimizations.")
+                    );
+                }
+            }
+        }
+        
+        #[cfg(target_os = "linux")]
+        {
+            let mem_output = Command::new("free").args(["-m"]).output();
+            if let Ok(output) = mem_output {
+                let content = String::from_utf8_lossy(&output.stdout);
+                for line in content.lines() {
+                    if line.starts_with("Mem:") {
+                        let parts: Vec<&str> = line.split_whitespace().collect();
+                        if parts.len() >= 4 {
+                            result["memory"] = json!({
+                                "total_mb": parts[1].parse::<u64>().unwrap_or(0),
+                                "used_mb": parts[2].parse::<u64>().unwrap_or(0),
+                                "free_mb": parts[3].parse::<u64>().unwrap_or(0),
+                                "unit": "MB",
+                            });
+                        }
+                    }
+                }
+            }
+        }
+        
+        // Get CPU cores
+        let nproc_output = Command::new("sysctl")
+            .args(["-n", "hw.ncpu"])
+            .output();
+        
+        if let Ok(output) = nproc_output {
+            let cores = String::from_utf8_lossy(&output.stdout).trim().parse().unwrap_or(0);
+            result["cpu_cores"] = json!(cores);
+        }
+    }
+    
+    Ok(json!({
+        "content": [{ "type": "text", "text": result.to_string() }],
+        "isError": false
+    }))
+}
+
+/// Validate MT5 configuration files
+pub async fn handle_validate_mt5_config(config: &Config, _args: &Value) -> Result<Value> {
+    let mt5_dir = config.mt5_dir()
+        .ok_or_else(|| anyhow::anyhow!("MT5 directory not configured"))?;
+    
+    let mut result = json!({
+        "terminal_ini": null,
+        "tester_ini": null,
+        "config_files_found": Vec::<String>::new(),
+        "errors": Vec::<String>::new(),
+        "warnings": Vec::<String>::new(),
+    });
+    
+    // Check terminal.ini
+    let terminal_ini = mt5_dir.join("terminal.ini");
+    if terminal_ini.exists() {
+        result["config_files_found"].as_array_mut().unwrap().push(json!("terminal.ini"));
+        
+        if let Ok(content) = fs::read_to_string(&terminal_ini) {
+            // Check for common issues
+            if !content.contains("[Common]") {
+                result["errors"].as_array_mut().unwrap().push(
+                    json!("terminal.ini missing [Common] section")
+                );
+            }
+            
+            // Extract key settings
+            let mut settings = serde_json::Map::new();
+            for line in content.lines() {
+                if line.starts_with("Login=") {
+                    settings.insert("login".to_string(), json!(line.trim_start_matches("Login=")));
+                } else if line.starts_with("Server=") {
+                    settings.insert("server".to_string(), json!(line.trim_start_matches("Server=")));
+                } else if line.starts_with("Expert=") {
+                    settings.insert("expert".to_string(), json!(line.trim_start_matches("Expert=")));
+                }
+            }
+            result["terminal_ini"] = json!(settings);
+        } else {
+            result["errors"].as_array_mut().unwrap().push(
+                json!("Could not read terminal.ini")
+            );
+        }
+    } else {
+        result["warnings"].as_array_mut().unwrap().push(
+            json!("terminal.ini not found")
+        );
+    }
+    
+    // Check for tester config
+    let tester_dir = mt5_dir.join("Tester");
+    if tester_dir.exists() {
+        if let Ok(entries) = fs::read_dir(&tester_dir) {
+            let ini_files: Vec<String> = entries
+                .flatten()
+                .filter_map(|e| {
+                    let p = e.path();
+                    if p.extension()?.to_str()? == "ini" {
+                        Some(p.file_name()?.to_string_lossy().to_string())
+                    } else {
+                        None
+                    }
+                })
+                .collect();
+            
+            if !ini_files.is_empty() {
+                result["tester_ini"] = json!(ini_files);
+            }
+        }
+    }
+    
+    // Check for common problems
+    let experts_dir = mt5_dir.join("MQL5").join("Experts");
+    if !experts_dir.exists() {
+        result["errors"].as_array_mut().unwrap().push(
+            json!("MQL5/Experts directory not found - MT5 installation may be incomplete")
+        );
+    }
+    
+    let has_errors = !result["errors"].as_array().unwrap().is_empty();
+    
+    Ok(json!({
+        "content": [{ "type": "text", "text": result.to_string() }],
+        "isError": has_errors
+    }))
+}
+
+/// Get Wine prefix detailed information
+pub async fn handle_get_wine_prefix_info(config: &Config, _args: &Value) -> Result<Value> {
+    let mt5_dir = config.mt5_dir()
+        .ok_or_else(|| anyhow::anyhow!("MT5 directory not configured"))?;
+    
+    let wine_prefix = mt5_dir
+        .parent()
+        .and_then(|p| p.parent())
+        .and_then(|p| p.parent());
+    
+    let mut result = json!({
+        "prefix_path": null,
+        "exists": false,
+        "windows_version": null,
+        "dll_overrides": Vec::<String>::new(),
+        "installed_programs": Vec::<String>::new(),
+        "registry_files": Vec::<String>::new(),
+        "drive_c_size_mb": 0,
+    });
+    
+    if let Some(prefix) = wine_prefix {
+        result["prefix_path"] = json!(prefix.to_string_lossy().to_string());
+        result["exists"] = json!(prefix.exists());
+        
+        if prefix.exists() {
+            // Check Windows version
+            let system_reg = prefix.join("system.reg");
+            if system_reg.exists() {
+                if let Ok(content) = fs::read_to_string(&system_reg) {
+                    for line in content.lines().take(50) {
+                        if line.contains("\"ProductName\"") {
+                            let parts: Vec<&str> = line.split('"').collect();
+                            if parts.len() >= 4 {
+                                result["windows_version"] = json!(parts[3]);
+                            }
+                        }
+                    }
+                }
+            }
+            
+            // Calculate drive_c size
+            let drive_c = prefix.join("drive_c");
+            if drive_c.exists() {
+                let mut size = 0u64;
+                fn calc_size(dir: &Path, total: &mut u64) {
+                    if let Ok(entries) = fs::read_dir(dir) {
+                        for e in entries.flatten() {
+                            let p = e.path();
+                            if p.is_file() {
+                                if let Ok(m) = e.metadata() {
+                                    *total += m.len();
+                                }
+                            } else if p.is_dir() {
+                                calc_size(&p, total);
+                            }
+                        }
+                    }
+                }
+                calc_size(&drive_c, &mut size);
+                result["drive_c_size_mb"] = json!((size / 1024 / 1024) as i64);
+            }
+            
+            // Check for installed programs
+            let prog_files = prefix.join("drive_c").join("Program Files");
+            if prog_files.exists() {
+                if let Ok(entries) = fs::read_dir(&prog_files) {
+                    let programs: Vec<String> = entries
+                        .flatten()
+                        .filter_map(|e| {
+                            let p = e.path();
+                            if p.is_dir() {
+                                Some(p.file_name()?.to_string_lossy().to_string())
+                            } else {
+                                None
+                            }
+                        })
+                        .collect();
+                    result["installed_programs"] = json!(programs);
+                }
+            }
+            
+            // List registry files
+            let reg_files: Vec<String> = vec![
+                "system.reg", "user.reg", "userdef.reg"
+            ]
+            .into_iter()
+            .filter(|f| prefix.join(f).exists())
+            .map(|f| f.to_string())
+            .collect();
+            result["registry_files"] = json!(reg_files);
+        }
+    }
+    
+    Ok(json!({
+        "content": [{ "type": "text", "text": result.to_string() }],
+        "isError": false
+    }))
+}
+
+/// Get backtest crash/failure information
+pub async fn handle_get_backtest_crash_info(config: &Config, args: &Value) -> Result<Value> {
+    let report_dir = args.get("report_dir").and_then(|v| v.as_str());
+    let check_recent = args.get("check_recent").and_then(|v| v.as_bool()).unwrap_or(true);
+    let hours_back = args.get("hours_back").and_then(|v| v.as_u64()).unwrap_or(6);
+    
+    let mut result = json!({
+        "crashes_found": Vec::<serde_json::Value>::new(),
+        "recent_failures": 0,
+        "common_patterns": Vec::<String>::new(),
+    });
+    
+    // Check specific report directory if provided
+    if let Some(dir) = report_dir {
+        let path = Path::new(dir);
+        if path.exists() {
+            // Check for incomplete markers
+            let incomplete_marker = path.join(".incomplete");
+            let error_log = path.join("error.log");
+            
+            if incomplete_marker.exists() {
+                result["crashes_found"].as_array_mut().unwrap().push(json!({
+                    "report_dir": dir,
+                    "type": "incomplete",
+                    "reason": "Backtest was interrupted or timed out",
+                }));
+            }
+            
+            if error_log.exists() {
+                if let Ok(content) = fs::read_to_string(&error_log) {
+                    result["crashes_found"].as_array_mut().unwrap().push(json!({
+                        "report_dir": dir,
+                        "type": "error_log",
+                        "content": content.lines().take(20).collect::<Vec<_>>().join("\n"),
+                    }));
+                }
+            }
+            
+            // Check if deals.csv is missing or empty
+            let deals_csv = path.join("deals.csv");
+            if !deals_csv.exists() {
+                result["crashes_found"].as_array_mut().unwrap().push(json!({
+                    "report_dir": dir,
+                    "type": "missing_deals",
+                    "reason": "deals.csv not found - backtest likely failed",
+                }));
+            } else if let Ok(meta) = deals_csv.metadata() {
+                if meta.len() < 100 {
+                    result["crashes_found"].as_array_mut().unwrap().push(json!({
+                        "report_dir": dir,
+                        "type": "empty_deals",
+                        "reason": "deals.csv is nearly empty - no trades were made",
+                    }));
+                }
+            }
+        }
+    }
+    
+    // Check recent reports if requested
+    if check_recent {
+        let reports_dir_str = config.get("reports_dir");
+        let reports_dir = Path::new(&reports_dir_str);
+        if reports_dir.exists() {
+            let cutoff = std::time::SystemTime::now() - std::time::Duration::from_secs(hours_back * 3600);
+            
+            if let Ok(entries) = fs::read_dir(&reports_dir) {
+                let mut failures = 0;
+                for entry in entries.flatten() {
+                    let path = entry.path();
+                    if path.is_dir() {
+                        if let Ok(meta) = entry.metadata() {
+                            if let Ok(modified) = meta.modified() {
+                                if modified >= cutoff {
+                                    // Check for failure indicators
+                                    let has_deals = path.join("deals.csv").exists();
+                                    let has_incomplete = path.join(".incomplete").exists();
+                                    
+                                    if !has_deals || has_incomplete {
+                                        failures += 1;
+                                    }
+                                }
+                            }
+                        }
+                    }
+                }
+                result["recent_failures"] = json!(failures);
+            }
+        }
+    }
+    
+    // Analyze common patterns
+    let crashes = result["crashes_found"].as_array().unwrap();
+    if !crashes.is_empty() {
+        let types: Vec<String> = crashes.iter()
+            .filter_map(|c| c.get("type").and_then(|v| v.as_str()).map(|s| s.to_string()))
+            .collect();
+        
+        if types.contains(&"missing_deals".to_string()) {
+            result["common_patterns"].as_array_mut().unwrap().push(
+                json!("Missing deals.csv suggests MT5 crashed during backtest")
+            );
+        }
+        if types.contains(&"incomplete".to_string()) {
+            result["common_patterns"].as_array_mut().unwrap().push(
+                json!("Incomplete markers indicate interruptions - check system resources")
+            );
+        }
+    }
+    
+    Ok(json!({
+        "content": [{ "type": "text", "text": result.to_string() }],
+        "isError": false
+    }))
+}
Author	SHA1	Message	Date
Devid HW	91af1bb6dd	release: v1.31.5 - Version bump 1.31.4 → 1.31.5 - server.json identifier URL updated (SHA256 set by CI after build) - CHANGELOG.md updated	2026-04-22 07:34:18 +07:00
Devid HW	bb6240657e	feat: add check_update and update tools with background auto-check On the first tool call of each session, a background task fires once and fetches the latest release tag from the GitHub API (5 s timeout). The result is cached in a static OnceLock for the lifetime of the process — no repeated network calls. check_update: returns cached result instantly, or fetches on demand if the background task hasn't completed yet. update: downloads the latest tarball for the current platform, extracts the binary, and atomically replaces the running executable via a temp-file rename. Requires MCP reconnect to activate new version. Platform support: macos-aarch64, macos-x86_64, linux-x86_64. Unsupported platforms return a build-from-source hint. Tool count: 87 → 89. README and MCP_TOOLS.md updated. Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>	2026-04-22 07:34:12 +07:00
Devid HW	6f773d6f09	release: v1.31.4 - Version bump 1.31.3 → 1.31.4 - server.json identifier URL updated (SHA256 set by CI after build) - CHANGELOG.md updated	2026-04-22 07:24:03 +07:00
Devid HW	c895eb9b33	fix: update all scripts for correctness and consistency release.sh: - Add --yes/-y flag (and CI=true) for non-interactive execution - Auto-resolve server.json conflict during rebase-before-push so CI's SHA256 commit no longer blocks the push - Fix major version bump (was double-assigned, now single expression) build-release.sh: - Remove cp of /CLAUDE.md (gitignored personal file — caused build failure) - Bundle all IDE setup docs (QUICKSTART, CLAUDE, CURSOR, VSCODE, WINDSURF) - Use set -euo pipefail; chmod +x binary; clean up dist/ subdir after tar build-rust.sh: - Replace Windsurf-specific install hint with generic cargo install --path . --force - Use set -euo pipefail setup.sh: - Replace "Antigravity" with "Claude Desktop" in comments and output Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>	2026-04-22 07:24:03 +07:00
github-actions[bot]	e0b1356afa	ci: update server.json SHA256 for v1.31.3 [skip ci]	2026-04-22 00:21:57 +00:00
Devid HW	23a0b7a375	release: v1.31.3 - Version bump 1.31.2 → 1.31.3 - server.json identifier URL updated (SHA256 set by CI after build) - CHANGELOG.md updated	2026-04-22 07:20:25 +07:00
Devid HW	ac5ec453d3	docs: clean up public repo — remove IDE files, fix stale refs - Untrack .claude/commands/ and .windsurf/workflows/ (personal IDE config) - Fix .gitignore: scope /CLAUDE.md to root only, add .claude/, remove incorrect .github/workflows/ exclusion (CI needs these tracked) - Rename docs/ANTIGRAVITY.md → docs/CLAUDE.md (Claude Desktop setup; "Antigravity" was an internal codename, confusing to public users) - Fix stale repo name: masdevid/mt5-mcp → masdevid/mt5-quant in README.md, docs/CLAUDE.md, CURSOR.md, QUICKSTART.md, VSCODE.md, WINDSURF.md Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>	2026-04-22 07:20:20 +07:00
github-actions[bot]	9f6c67d0ef	ci: update server.json SHA256 for v1.31.2 [skip ci]	2026-04-22 00:04:10 +00:00
Devid HW	262936a1d5	release: v1.31.2 - Version bump 1.31.1 → 1.31.2 - server.json identifier URL updated (SHA256 set by CI after build) - CHANGELOG.md updated	2026-04-22 07:02:18 +07:00
Devid HW	0ea1e779e0	refactor: reduce handler boilerplate in analysis and experts modules analysis.rs: extract required_str, ok_response, err_response, prepare_analysis, deal_to_json, is_closed_trade helpers — 15 granular handlers reduced from ~14 lines each to 4 lines each (-293 lines total). experts.rs: extract scan_mql_dir (replaces 6 identical WalkDir loops), BUILTIN_INDICATORS const (deduplicates list/search handlers), and copy_mql_to_project (merges copy_indicator/copy_script into one impl, public handlers delegate in 1 line each) (-256 lines total). Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>	2026-04-22 07:01:58 +07:00
github-actions[bot]	a94d75823a	ci: update server.json SHA256 for v1.31.1 [skip ci]	2026-04-21 23:32:59 +00:00
Devid HW	c7e1a03e1f	fix: registryType mcpbPackageType → mcpb MCP registry API error: 'unsupported registry type: mcpbPackageType' Schema examples show valid values: npm, pypi, oci, nuget, mcpb Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>	2026-04-22 06:31:16 +07:00
github-actions[bot]	23c67e80c5	ci: update server.json SHA256 for v1.31.1 [skip ci]	2026-04-21 23:29:49 +00:00
Devid HW	5db3c7c822	ci: fix mcp-publisher install — resolve URL from release assets API v1.6.0 dropped the version number from asset filenames (mcp-publisher_linux_amd64.tar.gz, not mcp-publisher_1.6.0_linux_amd64.tar.gz). Look up the linux/amd64 .tar.gz URL directly from the GitHub releases API instead of constructing it, so future renames don't break the workflow. Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>	2026-04-22 06:27:06 +07:00
github-actions[bot]	5eb753584d	ci: update server.json SHA256 for v1.31.1 [skip ci]	2026-04-21 23:18:49 +00:00
Devid HW	7c9eeef1c4	release: v1.31.1 - Version bump 1.31.0 → 1.31.1 - server.json identifier URL updated (SHA256 set by CI after build) - CHANGELOG.md updated	2026-04-22 06:15:56 +07:00
Devid HW	20a2d3d6e4	ci: add release automation scripts and workflows - scripts/release.sh: one-command local release (bump, changelog, tag, push) - .github/workflows/release.yml: fix SHA256 auto-commit back to master, dynamic mcp-publisher version detection, cargo cache, security hardening - .claude/commands/release.md: /release slash command - .windsurf/workflows/release.md: streamlined to use release.sh Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>	2026-04-22 06:15:38 +07:00
Devid HW	ec2a22ed30	ci: handle mcp-publisher registryType bug gracefully - Don't fail workflow if MCP publish fails - Add clear manual publish instructions - Reference web UI as alternative	2026-04-22 05:58:09 +07:00
Devid HW	94a8c42e68	fix: update registryType to mcpbPackageType - The registry expects mcpbPackageType not mcpb	2026-04-22 05:52:20 +07:00
Devid HW	7e97086909	ci: add checkout step to mcp-publish job - server.json is needed for mcp-publisher to work	2026-04-22 05:44:43 +07:00
Devid HW	7a8a8daeca	ci: use GitHub OIDC for MCP authentication - Replace non-existent --token-stdin with github-oidc method - Add id-token: write permission for OIDC - Remove MCP_PUBLISH_TOKEN (not needed with OIDC)	2026-04-22 05:39:46 +07:00
Devid HW	ba04405454	ci: fix mcp-publisher download URL - Change from wrong repo (mcp-publisher) to correct repo (registry) - Use proper tar.gz format with versioned release URL - Add tar extraction step - Update manual install instructions with correct URL	2026-04-22 05:32:55 +07:00
Devid HW	e40dbf63cf	ci: fix mcp-publisher download with error handling - Add curl -fsSL to fail on HTTP errors - Add graceful skip if download fails - Add manual publish instructions as fallback	2026-04-22 05:28:01 +07:00
Devid HW	0a2e076ad6	ci: rename release artifacts to mcp-mt5-quant-* format - Change mt5-quant-macos-arm64.tar.gz -> mcp-mt5-quant-macos-arm64.tar.gz - Change mt5-quant-linux-x64.tar.gz -> mcp-mt5-quant-linux-x64.tar.gz - Update build-mcp-package job extraction paths	2026-04-22 05:22:02 +07:00
Devid HW	8f71f1629d	docs: unify all docs to use single installation path ~/.local/bin/mt5-quant - Update WINDSURF.md: Use registry name io.github.masdevid/mt5-quant - Update CURSOR.md: Use ~/.local/bin/mt5-quant path - Update QUICKSTART.md: Use standard path for all clients - Update VSCODE.md: Use standard path - Update setup.sh: Prefer ~/.local/bin/mt5-quant over project path - Add fallback warning in setup.sh if project path used Single source of truth: ~/.local/bin/mt5-quant	2026-04-22 05:17:20 +07:00
Devid HW	ad17f9dba6	ci: add release-info job with Claude/Windsurf/Cursor commands - Add final job that displays copy-paste registration commands - Shows Claude Code, Windsurf, and VS Code/Cursor commands - Includes release URLs and success banner	2026-04-22 05:07:27 +07:00
Devid HW	33186d276a	ci: add automated MCP publishing and client registration docs - Update release workflow with MCP_PUBLISH_TOKEN for automated publishing - Remove OIDC requirement (using classic token) - Add MCP client registration commands for Windsurf, Claude, VS Code/Cursor - Add verification steps for post-release checking	2026-04-22 05:04:18 +07:00
Devid HW	3927343999	ci: update release workflow for manual triggers and MCP publish - Add workflow_dispatch with version input for manual releases - Remove tag-only condition - now works with manual triggers - Use github.event.inputs.version \|\| github.ref_name for flexibility - Add MCP package artifact upload - Add mcp-publish job with OIDC support instructions	2026-04-22 04:57:23 +07:00
Devid HW	6a5253186b	ci: add release workflow for automated builds - Build macOS and Linux binaries - Create MCP package with SHA256 - Upload to GitHub releases	2026-04-22 04:50:16 +07:00
Devid HW	76dac96583	docs: update release workflow to use single installation path - Use ~/.local/bin/mt5-quant as single installation location - Remove redundant project URL references - Simplify MCP registration for all clients (Claude, Windsurf, VS Code)	2026-04-22 04:48:56 +07:00
Devid HW	0bc410f613	feat: add Wine/MT5 debugging tools and update release workflow - Add 9 debugging/diagnostics tools for Wine/MT5 crash investigation - Update server.json with v1.30.0 and MCP package config - Update README.md with 85 tools count and debugging section - Update docs/MCP_TOOLS.md documentation - Enhance release workflow with MCP packaging job - Clean up mcp-package directory (now built in CI)	2026-04-22 04:41:41 +07:00
Devid HW	9be1296916	fix: shorten server.json description for MCP registry Description must be <=100 characters for registry validation. Previous: 96 chars -> Current: 82 chars	2026-04-20 03:04:17 +07:00
Devid HW	63c0470c6b	fix: update server.json with real SHA256 and shorten description - Description: 96 chars (was >100) - SHA256: calculated from local MCP package build	2026-04-20 02:54:28 +07:00
Devid HW	a1434914e9	chore: update release workflow and server.json for v1.30.0 - Update server.json: version 1.30.0, new download URL, updated description - Add MCP package build job to release workflow - Add SHA256 calculation for server.json - Update release workflow to upload mcp-* package	2026-04-20 02:51:36 +07:00
Devid HW	6ce8808948	v1.30.0: Add 10 Wine/MT5 debugging tools New debugging/diagnostics tools for crash investigation: - diagnose_wine: Check Wine installation and prefix health - get_mt5_logs: Get terminal/tester/metaeditor logs - search_mt5_errors: Search logs for error patterns - check_mt5_process: Check MT5 process status - kill_mt5_process: Kill stuck MT5 processes - check_system_resources: Check disk/memory/CPU - validate_mt5_config: Validate MT5 configuration - get_wine_prefix_info: Wine prefix details - get_backtest_crash_info: Investigate backtest failures Total tools: 85 Documentation updated in README.md and MCP_TOOLS.md	2026-04-20 02:25:07 +07:00
Devid HW	896aa6111e	feat: auto-detect and register MCP on multiple platforms - Add multi-platform MCP detection (Claude, Windsurf, Cursor, VS Code) - Auto-unregister previous installations before reinstall - Create platform-specific docs (CURSOR.md, VSCODE.md, ANTIGRAVITY.md) - Update Windsurf config to use correct JSON path (~/.codeium/windsurf/mcp_config.json) - Fix VS Code config format (uses 'servers' not 'mcpServers') - Remove MT5_MCP_HOME requirement from MCP configs (binary auto-detects) - Slim down README installation section with platform table - Update QUICKSTART.md with all platform instructions	2026-04-20 00:40:41 +07:00
Devid	aab34a1fde	Add GitHub Actions workflow for Rust project	2026-04-19 10:16:03 +07:00
Devid HW	96c7e0a9c2	chore: update server.json for v1.29.1 release	2026-04-19 10:10:55 +07:00
Devid HW	0c82501e18	chore: bump version to 1.29.1	2026-04-19 10:10:02 +07:00
Devid HW	2fef0db3c4	fix: wrap tools/list response in {tools: [...]} per MCP spec	2026-04-19 10:07:59 +07:00
Devid HW	95ceef3e3f	chore: update Cargo.lock for v1.29.0	2026-04-19 09:54:55 +07:00
Devid HW	662e9e324c	chore: update server.json for v1.29.0 release	2026-04-19 09:53:18 +07:00
Devid HW	6d1f940887	chore: bump version to 1.29.0	2026-04-19 09:52:15 +07:00
Devid HW	41c8b36826	fix: handle MCP notifications properly (notifications/initialized) - Add handle_notification() for notification messages (no id field) - Use CARGO_PKG_VERSION for dynamic version in initialize response - Skip serializing None fields in McpResponse and McpError - Remove verified field from setup info	2026-04-19 09:51:46 +07:00
Devid HW	42318b040a	docs: update README with Claude Code and Windsurf MCP installation instructions - Separate download/setup and MCP registration steps - Add detailed Claude Code registration with verification - Add Windsurf config.yaml examples (manual and automated) - Add absolute path requirement warning note	2026-04-19 09:45:47 +07:00
Devid HW	fdb45b8eed	fix: redirect tracing logs to stderr to keep stdout clean for MCP JSON tracing_subscriber::fmt::init() defaults to stdout, contaminating the MCP JSON-RPC stream and causing Claude Code and Windsurf to fail the initialize handshake with "failed to reconnect". Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>	2026-04-19 09:21:00 +07:00
Devid HW	b63bda64ab	chore: Update server.json for v1.28.0 release	2026-04-19 08:49:18 +07:00
Devid HW	cbed3d15af	docs: update README and MCP_TOOLS.md for 57 tools README.md: - Update tool count: 43 → 57 - Add Pre-flight & Validation section with get_active_account, check_symbol_data_status, check_mt5_status, validate_ea_syntax - Add Project Management section with init_project, create_set_template, export_report - Add History & Comparison section with get_backtest_history, compare_backtests - Add Search & Discovery section with search_experts, search_indicators, search_scripts, copy_indicator_to_project, copy_script_to_project MCP_TOOLS.md: - Update documentation status: 31 → 49 of 57 tools documented - Add 8 new utility tool schemas: - get_active_account - check_symbol_data_status - check_mt5_status - get_backtest_history - compare_backtests - init_project - validate_ea_syntax - create_set_template - export_report - Add workflow examples for pre-flight validation, project management, and history comparison	2026-04-19 08:28:07 +07:00
Devid HW	8cce92c409	chore: update server.json to v1.28.0 with new release URL and SHA256 Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>	2026-04-19 08:20:43 +07:00