diff --git a/docs/EA-041_Inside_Pullback_Trend/methodology.md b/docs/EA-041_Inside_Pullback_Trend/methodology.md
index 0e89199..daa1469 100644
--- a/docs/EA-041_Inside_Pullback_Trend/methodology.md
+++ b/docs/EA-041_Inside_Pullback_Trend/methodology.md
@@ -1,56 +1,889 @@
-# Documentation & Technical Specifications
+# XAUUSD MT5 EA Research Methodology
 
-Thư mục này chứa toàn bộ tài liệu kỹ thuật, quy trình phát triển, hướng dẫn vận hành và các tiêu chuẩn kiểm thử dành cho hệ thống Expert Advisor (EA).
+## 1. Purpose
+
+This document defines the standard research, development, backtesting, and validation methodology for all Expert Advisors in this repository.
+
+The objective is to develop XAUUSD trading strategies through a repeatable process:
+
+Research Idea
+→ Define Rules
+→ Implement EA
+→ Baseline Backtest
+→ Analyze
+→ Robustness Validation
+→ Forward Test
+→ Production Candidate
+
+The purpose is not to find the backtest with the highest historical profit.
+
+The purpose is to determine whether a trading hypothesis demonstrates sufficient evidence of a repeatable and robust edge.
 
 ---
 
-## 1. Cấu Trúc Thư Mục Docs
+# 2. Repository Structure
 
-```
-Docs/
-├── architecture/         # Thiết kế kiến trúc phần mềm & sơ đồ luồng (Flowchart)
-├── strategies/           # Mô tả chi tiết thuật toán & quy tắc vào/ra lệnh
-├── testing-standards/    # Tiêu chuẩn backtest & tối ưu hóa (Optimization)
-└── deployment/           # Hướng dẫn cài đặt, cấu hình & quản lý rủi ro trên VPS
-```
+Each strategy should follow the repository structure:
+
+xauusd-mt5-ea-research/
+│
+├── EAs/
+│   └── EA-XXX_Strategy_Name/
+│       ├── EA-XXX_Strategy_Name.mq5
+│       └── README.md
+│
+├── Backtest/
+│   └── EA-XXX_Strategy_Name/
+│       ├── README.md
+│       └── Strategy Tester artifacts
+│
+├── Research/
+│   └── README.md
+│
+├── docs/
+│   └── methodology.md
+│
+└── GitHub_Profile/
+    └── README.md
 
 ---
 
-## 2. Danh Mục Tài Liệu Chính
+# 3. Research Principles
 
-### Kiến Trúc & Mã Nguồn (`architecture/`)
+## 3.1 Hypothesis First
 
-* `mql5-framework-spec.md`: Cấu trúc Framework MQL5 tiêu chuẩn, quản lý thư viện và module lệnh.
-* `risk-management-module.md`: Cơ chế quản lý vốn, Stop Loss/Take Profit, Trailing Stop và bộ lọc Max Spread.
+Every EA must start from a clear market hypothesis.
 
-### Chi Tiết Thuật Toán (`strategies/`)
+Example:
 
-* `ema-crossover-logic.md`: Quy tắc kỹ thuật cho các chiến lược giao cắt đường trung bình động (EMA/SMA).
-* `inside-bar-pullback.md`: Logic nhận diện mô hình Inside Bar và xác nhận xu hướng chính.
+Trend
+→ Pullback / Consolidation
+→ Inside Bar
+→ Breakout
+→ Trend Continuation
 
-### Tiêu Chuẩn Kiểm Thử (`testing-standards/`)
+The strategy must explain why a market behavior could create an exploitable opportunity.
 
-* `backtest-checklist.md`: Quy trình thực hiện backtest với dữ liệu 100% Real Ticks.
-* `metrics-thresholds.md`: Bộ tiêu chí đánh giá EA:
-
-  * Profit Factor > 1.2
-  * Max Drawdown < 10%
-  * Sharpe Ratio > 1.0
-
-### Vận Hành & Triển Khai (`deployment/`)
-
-* `vps-setup-guide.md`: Hướng dẫn thiết lập môi trường MetaTrader 5 trên VPS.
-* `live-monitoring.md`: Quy trình theo dõi nhật ký lệnh (Logs) và xử lý sự cố kết nối.
+Do not begin with parameter optimization.
 
 ---
 
-## 3. Quy Chuẩn Đóng Góp Tài Liệu
+## 3.2 Rules Must Be Programmable
 
-1. Sử dụng định dạng **Markdown** chuẩn cho tất cả các tài liệu.
-2. Mọi thay đổi trong thuật toán EA phải được cập nhật tương ứng vào thư mục `strategies/`.
-3. Tài liệu mới cần bao gồm:
+Every trading idea must be converted into explicit rules that can be implemented in MQL5.
 
-   * Mục đích
-   * Thông số đầu vào (Inputs)
-   * Logic xử lý
-   * Ví dụ minh họa
+Avoid subjective rules such as:
+
+"Strong trend"
+
+"Good setup"
+
+"Beautiful Inside Bar"
+
+"Market looks bullish"
+
+Replace them with measurable definitions.
+
+Example:
+
+Price > EMA(50)
+
+Inside High <= Mother High
+
+Inside Low >= Mother Low
+
+Price > Mother High
+
+→ BUY
+
+---
+
+## 3.3 Simple Baseline First
+
+The first implementation should be the simplest version capable of testing the core hypothesis.
+
+Avoid adding unnecessary filters before the baseline is understood.
+
+Recommended sequence:
+
+Core Entry
++
+Basic SL
++
+Basic TP
++
+Basic Position Control
+
+Only after baseline testing should additional components be evaluated.
+
+Examples:
+
+Session Filter
+
+ATR Filter
+
+Mother Bar Filter
+
+News Filter
+
+Break Even
+
+Trailing Stop
+
+Dynamic Risk Management
+
+---
+
+# 4. EA Identification
+
+Every EA receives a unique identifier.
+
+Format:
+
+EA-XXX_Strategy_Name
+
+Example:
+
+EA-041_Inside_Pullback_Trend
+
+The identifier should remain stable throughout the research lifecycle.
+
+Do not reuse an existing EA number for a different strategy.
+
+---
+
+# 5. Strategy Documentation
+
+Each EA must contain:
+
+EAs/
+└── EA-XXX_Strategy_Name/
+    ├── EA-XXX_Strategy_Name.mq5
+    └── README.md
+
+The README should document at minimum:
+
+- Strategy objective
+- Market hypothesis
+- Entry logic
+- BUY conditions
+- SELL conditions
+- Exit logic
+- Stop Loss
+- Take Profit
+- Filters
+- Position management
+- Input parameters
+- Known limitations
+- Research status
+
+The README describes what the EA is designed to do.
+
+It must not claim profitability that has not been demonstrated by testing.
+
+---
+
+# 6. Implementation Method
+
+## Stage 1 — Define Strategy
+
+Convert the research idea into deterministic rules.
+
+Output:
+
+Strategy Specification
+
+---
+
+## Stage 2 — Implement EA
+
+Translate the rules into MQL5.
+
+Output:
+
+EA-XXX_Strategy_Name.mq5
+
+---
+
+## Stage 3 — Compile
+
+Compile using MetaEditor.
+
+Required result:
+
+0 Errors
+
+Warnings should be reviewed before testing.
+
+Compilation success does NOT validate trading logic.
+
+---
+
+## Stage 4 — Logic Verification
+
+Before large backtests, verify that the EA behaves as intended.
+
+Check:
+
+- Entry conditions
+- BUY / SELL direction
+- Stop Loss placement
+- Take Profit placement
+- Position count
+- Spread filter
+- New-bar logic
+- Magic Number
+- Trade comments
+- Position management
+
+Where practical, visually inspect trades in MT5 Strategy Tester.
+
+---
+
+# 7. Baseline Backtest
+
+Every EA must first receive a baseline test.
+
+The baseline exists to answer:
+
+"Does the simplest implementation show enough evidence to justify further research?"
+
+It is NOT an optimization test.
+
+Record:
+
+- EA version
+- Symbol
+- Broker symbol
+- Timeframe
+- Test period
+- Modeling / tick quality
+- Initial deposit
+- Leverage
+- Lot size
+- Spread constraints
+- SL
+- TP
+- Filters
+- Break Even state
+- Trailing state
+
+Never report a backtest result without preserving its configuration.
+
+---
+
+# 8. Backtest Evidence
+
+Backtest artifacts should be stored under:
+
+Backtest/
+└── EA-XXX_Strategy_Name/
+
+Preserve the original MetaTrader Strategy Tester report whenever possible.
+
+Recommended artifacts:
+
+README.md
+
+Strategy Tester HTML report
+
+Balance / Equity graph
+
+Trade distribution charts
+
+MFE / MAE chart
+
+Holding-time chart
+
+Additional robustness reports
+
+The README summarizes the result.
+
+The original Strategy Tester report remains the primary evidence.
+
+---
+
+# 9. Core Metrics
+
+At minimum, evaluate:
+
+## Profitability
+
+- Total Net Profit
+- Gross Profit
+- Gross Loss
+- Profit Factor
+- Expected Payoff
+
+## Risk
+
+- Balance Drawdown
+- Equity Drawdown
+- Maximum Drawdown
+- Recovery Factor
+
+## Trade Statistics
+
+- Total Trades
+- Winning Trades
+- Losing Trades
+- Win Rate
+- Average Winner
+- Average Loser
+- Largest Winner
+- Largest Loser
+
+## Sequence Risk
+
+- Maximum Consecutive Wins
+- Maximum Consecutive Losses
+- Average Consecutive Wins
+- Average Consecutive Losses
+
+## Trade Behavior
+
+Where available:
+
+- MFE
+- MAE
+- Holding Time
+- LONG performance
+- SHORT performance
+
+No single metric determines strategy quality.
+
+Metrics must be evaluated together.
+
+---
+
+# 10. Trade Sample Size
+
+Small samples can produce misleading results.
+
+A positive result with only a small number of trades should be treated as preliminary evidence.
+
+Example:
+
+57 trades
++
+Profit Factor 1.21
+
+may justify continued research.
+
+It does NOT establish a robust trading edge.
+
+Priority should be given to increasing the number of independent observations before aggressive optimization.
+
+---
+
+# 11. Historical Coverage
+
+After baseline testing, extend the historical period.
+
+Where reliable data is available, research should include multiple market regimes.
+
+Examples:
+
+Trending markets
+
+Range-bound markets
+
+High-volatility periods
+
+Low-volatility periods
+
+Bullish Gold regimes
+
+Bearish / corrective Gold regimes
+
+The objective is to determine whether performance depends excessively on one specific market environment.
+
+---
+
+# 12. Multi-Timeframe Testing
+
+Strategies should be tested on reasonable neighboring timeframes where the logic remains meaningful.
+
+Typical XAUUSD research set:
+
+M1
+
+M5
+
+M15
+
+H1
+
+A strategy does not need identical performance across all timeframes.
+
+However, extreme dependence on one isolated timeframe should be investigated.
+
+---
+
+# 13. Parameter Research
+
+Parameter research begins only after the baseline demonstrates enough evidence to continue.
+
+Possible parameters include:
+
+EMA Period
+
+Stop Loss
+
+Take Profit
+
+ATR Threshold
+
+Pattern Size
+
+Session
+
+Spread
+
+Break Even
+
+Trailing Stop
+
+Do not optimize all parameters simultaneously in the first iteration.
+
+Change one logical component or a small related parameter group at a time.
+
+---
+
+# 14. Parameter Robustness
+
+Do not select a parameter solely because it produces the highest historical profit.
+
+Prefer parameter regions where neighboring values also perform reasonably.
+
+Example:
+
+If:
+
+Parameter 49 → poor
+
+Parameter 50 → exceptional
+
+Parameter 51 → poor
+
+then Parameter 50 may represent overfitting.
+
+A more credible region resembles:
+
+Parameter 45 → acceptable
+
+Parameter 50 → good
+
+Parameter 55 → acceptable
+
+The objective is stability, not a single historical optimum.
+
+---
+
+# 15. In-Sample and Out-of-Sample
+
+After promising parameters are identified, separate research data from validation data.
+
+## In-Sample
+
+Used for:
+
+Research
+
+Parameter development
+
+Strategy improvement
+
+## Out-of-Sample
+
+Used for:
+
+Independent validation
+
+The Out-of-Sample period must not be repeatedly optimized against.
+
+Otherwise it effectively becomes part of the In-Sample dataset.
+
+---
+
+# 16. Walk-Forward Testing
+
+Promising strategies should eventually undergo walk-forward testing.
+
+Concept:
+
+Train / Research
+→ Test
+
+Move Window
+
+Train / Research
+→ Test
+
+Move Window
+
+Repeat
+
+The objective is to determine whether parameters selected from historical information continue to perform on unseen subsequent data.
+
+---
+
+# 17. Monte Carlo / Robustness Testing
+
+Strategies that survive Out-of-Sample testing should undergo additional robustness analysis where practical.
+
+Possible tests:
+
+- Trade-order randomization
+- Spread variation
+- Slippage variation
+- Execution variation
+- Parameter perturbation
+- Trade removal
+- Return-sequence simulation
+
+The purpose is to estimate how fragile the observed backtest result may be.
+
+---
+
+# 18. Transaction Costs
+
+XAUUSD backtests must account for realistic trading conditions.
+
+Important factors include:
+
+Spread
+
+Commission
+
+Slippage
+
+Execution
+
+Broker symbol specifications
+
+Swap where relevant
+
+A strategy that only works under unrealistically favorable trading costs should not be considered validated.
+
+---
+
+# 19. Broker Dependency
+
+XAUUSD specifications can differ between brokers.
+
+Record:
+
+Broker
+
+Symbol
+
+Digits
+
+Point size
+
+Contract specifications
+
+Spread behavior
+
+Commission
+
+Leverage
+
+Execution environment
+
+Example:
+
+XAUUSD
+
+and
+
+XAUUSD.PRO
+
+must not automatically be assumed to behave identically.
+
+---
+
+# 20. Risk Management Research
+
+Fixed lot size should generally be used during early strategy research when the objective is to study the trading logic itself.
+
+Example:
+
+Lot = 0.01
+
+This helps separate:
+
+Strategy Edge
+
+from
+
+Position Sizing Effects
+
+Dynamic risk sizing may be introduced later.
+
+Examples:
+
+Fixed % Risk
+
+Volatility-based sizing
+
+ATR-based sizing
+
+Do not use aggressive money management to make a weak strategy appear profitable.
+
+---
+
+# 21. Anti-Overfitting Rules
+
+Avoid:
+
+Testing hundreds of combinations and selecting only the best result.
+
+Adding filters solely because they improve historical profit.
+
+Removing losing periods without a market-based reason.
+
+Selecting one unusually profitable timeframe.
+
+Repeatedly optimizing against the Out-of-Sample period.
+
+Changing strategy rules after every losing backtest.
+
+Prefer:
+
+Simple hypotheses.
+
+Few parameters.
+
+Economic / market rationale.
+
+Large samples.
+
+Stable parameter regions.
+
+Independent validation.
+
+---
+
+# 22. Research Decision States
+
+Each EA should have a clear research state.
+
+Recommended states:
+
+IDEA
+
+IMPLEMENTED
+
+BASELINE TESTED
+
+RESEARCH IN PROGRESS
+
+ROBUSTNESS TESTING
+
+FORWARD TESTING
+
+PRODUCTION CANDIDATE
+
+REJECTED
+
+A positive backtest does not automatically move an EA to PRODUCTION CANDIDATE.
+
+---
+
+# 23. PASS / FAIL Philosophy
+
+Testing should answer specific questions.
+
+Example:
+
+Question:
+
+Does the baseline strategy produce enough evidence to justify further research?
+
+Possible decision:
+
+PASS FOR FURTHER RESEARCH
+
+This does NOT mean:
+
+PASS FOR LIVE TRADING
+
+Different validation stages require different evidence.
+
+---
+
+# 24. Evidence Requirement
+
+A research stage should not be considered complete without evidence.
+
+Evidence may include:
+
+- Source code
+- Compilation result
+- Strategy Tester report
+- Charts
+- Parameter configuration
+- Research notes
+- Out-of-Sample results
+- Walk-forward results
+- Forward-test records
+
+Claims should be traceable to stored artifacts.
+
+---
+
+# 25. Reproducibility
+
+Another researcher should be able to reproduce the test using:
+
+EA source code
+
++
+
+Input parameters
+
++
+
+Symbol
+
++
+
+Timeframe
+
++
+
+Historical period
+
++
+
+Broker/test environment
+
++
+
+Testing assumptions
+
+If the result cannot reasonably be reproduced, it should not be treated as strong evidence.
+
+---
+
+# 26. Research Workflow
+
+Standard workflow:
+
+IDEA
+↓
+MARKET HYPOTHESIS
+↓
+PROGRAMMABLE RULES
+↓
+MQL5 IMPLEMENTATION
+↓
+COMPILE
+↓
+LOGIC CHECK
+↓
+BASELINE BACKTEST
+↓
+RESULT ANALYSIS
+↓
+LONG-HISTORY TEST
+↓
+MULTI-TIMEFRAME TEST
+↓
+FILTER RESEARCH
+↓
+PARAMETER ROBUSTNESS
+↓
+OUT-OF-SAMPLE
+↓
+WALK-FORWARD
+↓
+MONTE CARLO / ROBUSTNESS
+↓
+FORWARD TEST
+↓
+PRODUCTION CANDIDATE
+
+At any stage:
+
+Insufficient Evidence
+or
+Failed Robustness
+
+→ REVISE or REJECT
+
+---
+
+# 27. Current Example — EA-041
+
+EA-041_Inside_Pullback_Trend currently provides the first implementation example for this methodology.
+
+Current completed stages:
+
+Strategy Defined
+
+MQL5 Implementation
+
+Baseline Backtest
+
+Initial Research Analysis
+
+Current baseline:
+
+Symbol:
+XAUUSD.PRO
+
+Timeframe:
+M1
+
+Period:
+2026-01-02 → 2026-09-03
+
+Net Profit:
++$26.78
+
+Profit Factor:
+1.21
+
+Maximum Equity Drawdown:
+4.50%
+
+Total Trades:
+57
+
+Current interpretation:
+
+BASELINE COMPLETED
+
+RESEARCH CONTINUES
+
+NOT APPROVED FOR LIVE TRADING
+
+The next priority is increasing historical coverage and comparing multiple timeframes before deeper optimization.
+
+---
+
+# 28. Final Principle
+
+The repository should optimize for:
+
+REPRODUCIBLE EVIDENCE
+
+not:
+
+IMPRESSIVE BACKTEST SCREENSHOTS
+
+A strong EA should survive progressively harder validation.
+
+The research process therefore follows:
+
+Simple
+→ Test
+→ Measure
+→ Validate
+→ Improve
+→ Retest
+
+Only strategies that continue to demonstrate acceptable behavior through independent validation should advance toward live evaluation.