A6-9V/MQL5-Google-Onedrive
1
0
Fork 0
forked from LengKundee/MQL5-Google-Onedrive
Code Pull requests Activity
https://github.com/A6-9V/my-drive-projects.git @Jules is an ai agent from google that manage the trading system for this repository https://studio.firebase.google.com/studio-6309792933
42 commits 269 branches 0 tags 331 MiB
Find a file
Exact
google-labs-jules[bot] 4ca7107987 ⚡ Bolt: Cache MTF indicator results to avoid redundant calculations 
💡 What:
This change introduces a caching mechanism for the multi-timeframe (MTF) trend confirmation logic in the `GetMTFDir()` function. The result of the EMA calculation is now stored and only recalculated when a new bar forms on the lower timeframe.

🎯 Why:
The `GetMTFDir()` function was being called on every `OnTick` event, which can fire multiple times per second. This led to expensive `CopyBuffer` calls to retrieve EMA values from a lower timeframe, even when those values had not changed.

📊 Impact:
This optimization dramatically reduces the number of expensive `CopyBuffer` calls. Instead of running on every tick, the calculation now only runs once per bar on the lower timeframe. For an H1 chart using M5 confirmation, this reduces calculations from potentially hundreds per minute to just one every 5 minutes, significantly improving the EA's performance and reducing CPU load.

🔬 Measurement:
The performance improvement can be verified by adding counters to the `GetMTFDir` function. A counter incremented at the start of the function will show a high call rate, while a counter incremented only when the recalculation logic is triggered will show a much lower rate, confirming the caching is effective.
2026-01-02 17:04:32 +00:00
.github feat: Conditionally enable OneDrive sync based on config 2025-12-27 06:22:19 +00:00
docs Docs: Add GitHub org and private repo access instructions 2025-12-28 04:16:00 +00:00
mt5/MQL5 ⚡ Bolt: Cache MTF indicator results to avoid redundant calculations 2026-01-02 17:04:32 +00:00
scripts feat: Add GitHub Actions for CI, automerge, and sync 2025-12-27 06:02:09 +00:00
.editorconfig feat: Add GitHub Actions for CI, automerge, and sync 2025-12-27 06:02:09 +00:00
.gitignore feat: Add MT5 packaging and deployment scripts 2025-12-26 08:26:52 +00:00
LICENSE Initial commit 2025-12-26 13:13:16 +07:00
README.md Add render workspace link to README 2025-12-28 05:31:17 +00:00

README.md

SMC + Trend Breakout (MTF) for Exness MT5

This repo contains:

  • mt5/MQL5/Indicators/SMC_TrendBreakout_MTF.mq5: visual indicator (BOS/CHoCH + Donchian breakout + lower-timeframe confirmation).
  • mt5/MQL5/Experts/SMC_TrendBreakout_MTF_EA.mq5: Expert Advisor (alerts + optional auto-trading).

Render workspace

My Blue watermelon Workspace
tea-d1joqqi4d50c738aiujg

Install into Exness MetaTrader 5

  1. Open Exness MT5.
  2. Go to File → Open Data Folder.
  3. Copy:
    • SMC_TrendBreakout_MTF.mq5 to MQL5/Indicators/
    • SMC_TrendBreakout_MTF_EA.mq5 to MQL5/Experts/
  4. In MT5, open MetaEditor (or press F4) and compile the files.
  5. Back in MT5: Navigator → Refresh.

Optional: package / deploy helpers

  • Create a zip you can copy to your PC:
    • bash scripts/package_mt5.sh → outputs dist/Exness_MT5_MQL5.zip
  • Copy directly into your MT5 Data Folder (run this on the machine that has MT5 installed):
    • bash scripts/deploy_mt5.sh "/path/from/MT5/File->Open Data Folder"

GitHub automation (reviews, CI, auto-merge, OneDrive sync)

This repo includes GitHub Actions workflows under .github/workflows/:

  • CI (CI): runs on pull requests and pushes to main/master
    • Validates repo structure
    • Builds dist/Exness_MT5_MQL5.zip and uploads it as an artifact
  • Auto-merge enablement (Enable auto-merge (label-driven)): if a PR has the label automerge, it will enable GitHub’s auto-merge (squash). Your branch protection rules still control when it can merge (required reviews, required CI, etc.).
  • OneDrive sync (Sync to OneDrive (rclone)): on pushes to main (and manual runs), syncs mt5/MQL5 to OneDrive via rclone.

Recommended repo settings (GitHub → Settings):

  • Branch protection (main):
    • Require pull request reviews (at least 1)
    • Require status checks: CI / validate-and-package
    • (Optional) Require CODEOWNERS review
  • Auto-merge: enable “Allow auto-merge” in repo settings

OneDrive sync setup (required secrets):

  • RCLONE_CONFIG_B64: base64 of your rclone.conf containing a OneDrive remote.

Example (run locally, then paste into GitHub Secrets):

rclone config
base64 -w0 ~/.config/rclone/rclone.conf

Optional secrets:

  • ONEDRIVE_REMOTE: remote name in rclone.conf (default: onedrive)
  • ONEDRIVE_PATH: destination folder path (default: Apps/MT5/MQL5)

Use the indicator

  • Attach SMC_TrendBreakout_MTF to a chart (your main timeframe).
  • Set LowerTF to a smaller timeframe (ex: main = M15, lower = M5 or M1).
  • Signals require lower-TF confirmation by default (EMA fast/slow direction).

Use the EA (push to terminal + optional auto trading)

  • Attach SMC_TrendBreakout_MTF_EA to a chart.
  • Enable Algo Trading in MT5 if you want auto entries.
  • If you want phone push alerts:
    • MT5 → Tools → Options → Notifications
    • enable push notifications and set your MetaQuotes ID.

Auto SL/TP + risk management (EA)

In SMC_TrendBreakout_MTF_EA:

  • SLMode
    • SL_ATR: SL = ATR × ATR_SL_Mult
    • SL_SWING: SL beyond last confirmed fractal swing (with SwingSLBufferPoints), fallback to ATR if swing is missing/invalid
    • SL_FIXED_POINTS: SL = FixedSLPoints
  • TPMode
    • TP_RR: TP = RR × SL distance
    • TP_FIXED_POINTS: TP = FixedTPPoints
    • TP_DONCHIAN_WIDTH: TP = Donchian channel width × DonchianTP_Mult (fallback to ATR width if needed)
  • RiskPercent
    • If RiskPercent > 0, lots are calculated from SL distance so the money at risk ≈ RiskPercent of Equity (or Balance if you disable RiskUseEquity).
    • RiskClampToFreeMargin can reduce lots if required margin is too high.

Notes / safety

  • This is a rules-based implementation of common “SMC” ideas (fractal swing BOS/CHoCH) and a Donchian breakout.
  • Test in Strategy Tester and/or demo before using real funds.

Project links

Contact