strategy-backtest
Skill58 starsChanged 3 months agoMITView on GitHub
Compare strategies on an IDX symbol, backtest the survivor walk-forward, and hand the result to TradingView as Pine Script. Use when the user asks whether a strategy works, to test an idea, or for a Pine script.
2 other projects use a version of this.
Adopted by
Most recently taken up by
- Weaxs/stock-analysis-plugintheir own version, first seen Nov 15, 2026
- BryceYuuu/open-workflow-skillstheir own version, first seen Sep 5, 2026
Other versions2 projects use a version of this.
Word for word · 0
None found.
Edited · 0
None found.
Wrote their own version · 2
- Weaxs/stock-analysis-plugin★ 95same name
- BryceYuuu/open-workflow-skills★ 1same name
How matches are found. Word for word: the exact same text. Edited: shares passages with this file (“99% same” means 99% of its five-word phrases also appear here; 10% or more counts). Their own version: similar meaning, or, weakest, the same name. We can’t tell which project had it first. Projects are counted among the 200 most-starred candidates, and mentions are links to the repository we found, up to 50 per site, so both counts are a floor.
In this repository9 other files it gives its agents.
- CLAUDE.mdCLAUDE.md
- bandar-checkSkill
- chart-markupSkill
- morning-scanSkill
- stockbit-authSkill
- stockbit-statusSkill
- stock-deep-diveSkill
- trade-with-guardrailsSkill
- watchSkill
The file
Files
Strategy backtest
The sequence
strategy_compare symbol=…— nine built-in strategies over the same bars, ranked. This is the cheap first pass: it tells you whether anything on this symbol has an edge worth testing properly, and the bars are fetched once for all of them.backtest symbol=… strategy=…on the one or two that stood out. Setwalk_forward=trueand give itfolds. An in-sample result on two years of daily bars is a description of the past, and presenting one as a strategy is the single most misleading thing this server can be used to do.- Set
commission_buy_pctandcommission_sell_pctto the user's real fees. IDX defaults are roughly 0.15% buy and 0.25% sell, and the difference between gross and net is where most paper edges live. stop_loss_pct,take_profit_pctandmax_hold_barschange the answer more than the entry rule usually does.
- Set
pine_script symbol=… kind=strategyto hand it to TradingView, with the entry and exit conditions carried across.
workflow_run name=strategy_check and name=pine_handoff package steps 1–2 and 3.
What to report, and what not to
warningsandinconclusiveare the result, not a footnote. Read them out. If the backtest says the sample was too small, that is the answer to the user's question.- ARA/ARB-locked bars break fills. A limit-up day has no liquidity at the printed price. The backtester flags these; a strategy whose returns come from limit-locked entries did not happen.
- Two years of daily bars is about 480 sessions. That supports claims about daily setups. It does not support a claim about monthly seasonality, and it does not survive being sliced by regime.
- Trade count matters more than return. Nine trades with a 60% win rate is not a 60% win rate.
- Never annualise a backtest or extrapolate it forward.
Presenting it
Give the net return after commission, the trade count, the maximum drawdown, and the walk-forward folds side by side — if fold three lost money, that belongs in the first paragraph. Then the warnings. Then, plainly: backtested results do not predict future returns, and this is not advice.
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

