Audit fixes: public repo readiness + MCP improvements

- Add MIT LICENSE
- Anonymize broker path in config.example.yaml; add slippage_points + risk_per_trade
- Fix encoding ansi→cp1252 in signals.py (cross-platform safe)
- Fix install.py silently ignoring pip errors
- Cap get_ohlcv at 2000 bars; add ping tool; clarify HTML report UX in descriptions
- Escape symbol/timeframe in HTML report to prevent malformed output
- Move run_*.py to scripts/; move example CSVs to examples/
- Update .gitignore to block accidental CSV commits from root
- README: Python 3.9+ requirement, Chart.js internet note, new sections for examples/ and scripts/

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

.gitignore

@@ -9,3 +9,8 @@ __pycache__/
 *.pyc
 .venv/
 venv/
+
+# signal CSVs — example files in examples/ are tracked; any CSV dropped in
+# the root or scripts/ is personal data and should not be committed
+/*.csv
+/scripts/*.csv

LICENSE

@@ -0,0 +1,21 @@
+MIT License
+
+Copyright (c) 2026 davidcyberdash
+
+Permission is hereby granted, free of charge, to any person obtaining a copy
+of this software and associated documentation files (the "Software"), to deal
+in the Software without restriction, including without limitation the rights
+to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
+copies of the Software, and to permit persons to whom the Software is
+furnished to do so, subject to the following conditions:
+
+The above copyright notice and this permission notice shall be included in all
+copies or substantial portions of the Software.
+
+THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+SOFTWARE.

README.md

@@ -2,6 +2,8 @@
 
 **Backtest any MetaTrader 5 indicator by talking to Claude.**
 
+**Requirements:** Python 3.9+ · MetaTrader 5 (Windows) · Claude Code
+
 Drop one include into your indicator, and MBT lets you fetch live data, read
 your indicator's signals, and run a full backtest with an HTML report — all
 through Claude, no scripting.
@@ -129,8 +131,31 @@ a timestamp.
 
 | Tool | Purpose |
 |------|---------|
-| `get_ohlcv` | live OHLCV bars for any symbol/timeframe |
+| `ping` | check MT5 is running and reachable |
+| `get_ohlcv` | live OHLCV bars for any symbol/timeframe (max 2000) |
 | `get_signals` | read your indicator's logged signals |
-| `backtest` | replay + full metrics + HTML report |
+| `backtest` | replay + full metrics + HTML report (requires internet for chart) |
 | `validate_signals` | check SL/TP geometry of every signal |
 | `get_config` | show the active terminal + signal file |
+
+> **Note:** the HTML report is written to `reports/` on your machine and opened in a
+> browser. Claude cannot open it directly — it will give you the file path.
+
+---
+
+## Examples
+
+`examples/` contains sample signal CSVs you can use to try MBT without attaching
+an indicator. Point `signal_file` in `config.yaml` to any of them:
+
+```yaml
+signal_file: "C:/abs/path/to/MBT/examples/test_signals.csv"
+```
+
+---
+
+## Research scripts
+
+`scripts/` contains one-off analysis scripts used during strategy development.
+They are not part of the core toolkit and have no documentation — treat them as
+reference material, not user tools.

config.example.yaml

@@ -3,7 +3,7 @@
 
 # Full path to the terminal64.exe of the MT5 instance your indicator runs on.
 # Right-click your MT5 shortcut -> Properties -> Target to find it.
-mt5_path: "C:/Users/Administrator/Desktop/Ava Trade MT5 Terminal/terminal64.exe"
+mt5_path: "C:/Path/To/Your/terminal64.exe"
 
 # The signal file your indicator writes (via SignalLogger.mqh).
 # Either a bare filename (looked up inside the MT5 .../MQL5/Files folder)
@@ -22,3 +22,11 @@ ambiguous_bar: "loss"
 
 # Where generated HTML reports are written (relative to the MBT folder).
 reports_dir: "reports"
+
+# Spread + slippage allowance in points, used to estimate transaction cost per trade.
+# 5 is a reasonable default for most forex pairs; increase for gold/indices.
+slippage_points: 5.0
+
+# Risk per trade as a decimal (0.01 = 1%) — used only to translate R units into
+# account-percentage returns in the backtest report. Does not affect WIN/LOSS outcomes.
+risk_per_trade: 0.01

core/report_html.py

@@ -5,6 +5,7 @@ Uses Chart.js from a CDN so the output is a single self-contained .html the
 user can open in any browser or screenshot for content.
 """
 
+import html as _html
 import os
 from datetime import datetime
 
@@ -33,6 +34,8 @@ def render(rep: BacktestReport, filename: str = None) -> str:
 
     path = os.path.join(reports_dir(), filename)
 
+    sym = _html.escape(rep.symbol)
+    tf  = _html.escape(rep.timeframe)
     pf_str = "∞" if rep.profit_factor == float("inf") else f"{rep.profit_factor:.2f}"
 
     cards = "".join([
@@ -79,7 +82,7 @@ def render(rep: BacktestReport, filename: str = None) -> str:
 <html lang="en">
 <head>
 <meta charset="utf-8">
-<title>MBT Backtest — {rep.symbol} {rep.timeframe}</title>
+<title>MBT Backtest — {sym} {tf}</title>
 <script src="https://cdn.jsdelivr.net/npm/chart.js"></script>
 <style>
   body {{ background:#0d1117; color:#e6edf3; font-family:-apple-system,Segoe UI,Roboto,sans-serif;
@@ -102,7 +105,7 @@ def render(rep: BacktestReport, filename: str = None) -> str:
 </style>
 </head>
 <body>
-  <h1>Backtest Report — {rep.symbol} {rep.timeframe}</h1>
+  <h1>Backtest Report — {sym} {tf}</h1>
   <div class="sub">{rep.total} trades taken · {rep.signals_skipped} signals skipped (trade already open)
       · generated {datetime.now().strftime('%Y-%m-%d %H:%M')}
       · ambiguous bars counted as losses (conservative)</div>

core/signals.py

@@ -62,7 +62,7 @@ def load_signals(path: Optional[str] = None) -> List[Signal]:
     cfg  = load_config()
     path = path or signal_file_path()
 
-    with open(path, encoding="ansi", newline="") as f:
+    with open(path, encoding="cp1252", errors="replace", newline="") as f:
         rows = list(csv.reader(f))
 
     if not rows:

install.py

@@ -25,8 +25,10 @@ def step(msg):
 
 def install_deps():
     step("Installing Python dependencies")
-    subprocess.run([sys.executable, "-m", "pip", "install", "-r",
-                    os.path.join(ROOT, "requirements.txt")], check=False)
+    result = subprocess.run([sys.executable, "-m", "pip", "install", "-r",
+                             os.path.join(ROOT, "requirements.txt")])
+    if result.returncode != 0:
+        print("WARNING: pip install failed. Check the error above before continuing.")
 
 
 def make_config():

mcp_server.py

@@ -15,7 +15,7 @@ sys.path.insert(0, os.path.dirname(os.path.abspath(__file__)))
 
 from mcp.server.fastmcp import FastMCP
 
-from core.connection import load_config, signal_file_path
+from core.connection import load_config, connect, signal_file_path
 from core.ohlcv      import fetch_recent
 from core.signals    import load_signals
 from core.backtest   import run_backtest, report_to_dict
@@ -29,7 +29,9 @@ def get_ohlcv(symbol: str, timeframe: str, count: int = 100) -> dict:
     """
     Fetch live OHLCV bars from the configured MT5 terminal (newest first).
     timeframe: 1m 5m 15m 30m 1h 4h 1d 1w.  Generic — works for any symbol.
+    count: max 2000 bars.
     """
+    count = min(count, 2000)
     try:
         bars = fetch_recent(symbol, timeframe, count)
         return {"symbol": symbol, "timeframe": timeframe, "count": len(bars), "bars": bars}
@@ -75,7 +77,9 @@ def backtest(since_date: str = None, html_report: bool = True) -> dict:
     profit factor, expectancy, drawdown, per-regime breakdown) in R units.
 
     since_date:   optional 'YYYY-MM-DD' filter on signals.
-    html_report:  also write a standalone HTML report with an equity curve.
+    html_report:  write a standalone HTML report to reports/ and return its path.
+                  The report requires an internet connection (loads Chart.js from CDN).
+                  Claude cannot open the file — share the path with the user to open in a browser.
     """
     try:
         sigs = load_signals()
@@ -124,7 +128,8 @@ def validate_signals() -> dict:
 
 @mcp.tool()
 def get_config() -> dict:
-    """Show the active MBT configuration (which terminal and signal file are in use)."""
+    """Show the active MBT configuration (which terminal and signal file are in use).
+    Note: the output includes local file paths — do not share it publicly."""
     try:
         cfg = load_config()
         return {"config": cfg, "resolved_signal_file": signal_file_path()}
@@ -132,5 +137,25 @@ def get_config() -> dict:
         return {"error": str(e)}
 
 
+@mcp.tool()
+def ping() -> dict:
+    """Check whether the MT5 terminal is running and reachable. Use this first
+    if any other tool returns an error, to confirm the connection is alive."""
+    try:
+        connect()
+        import MetaTrader5 as mt5
+        info = mt5.terminal_info()
+        if info is None:
+            return {"connected": False, "error": "terminal_info() returned None"}
+        return {
+            "connected": True,
+            "terminal": info.name,
+            "build": info.build,
+            "trade_allowed": info.trade_allowed,
+        }
+    except Exception as e:
+        return {"connected": False, "error": str(e)}
+
+
 if __name__ == "__main__":
     mcp.run(transport="stdio")