Search the docs, videos, blog, apps, teams, skills and release notes.
At a glance
What it is
Head-to-head strategy comparison, snapshot state, sweep params on strategy A, restore, sweep strategy B, then compare metrics side-by-side.
A skill is a written procedure a bot loads when a job calls for it. This one is a single file, SKILL.md, and the whole file is on this page.
The whole file, exactly as it downloads. Read it before you rely on it.
SKILL.md
103 lines · 603 words
You are running two Pine strategies head-to-head on the same symbol/timeframe and reporting a data-backed winner.
The comparison is only valid if both strategies run on the same chart state.
chart_get_state — confirm symbol, timeframe, and chart type
state_snapshot
Head-to-head strategy comparison, snapshot state, sweep params on strategy A, restore, sweep strategy B, then compare metrics side-by-side.
Nothing. It ships with Murage, which is open source. Your bots run on the AI plan you already have.
Yes. Open it in Settings → Skills to read, edit or duplicate it. Skill Guard scans it again after every edit.
Download the app, connect the AI you already pay for, and tell your Chief of Staff what needs doing. Plan on about ten minutes from install to a working team.
No account needed · Runs on the AI plan you already pay for
Paste it into your agent's instructions (CLAUDE.md, AGENTS.md or a custom GPT), or add it to a bot in Murage.
Ask Ember for the bot, then switch the skill on in its settings.
Step 1
Send your Chief of Staff this message. It proposes the bot and its role first, and creates it when you say go.
Send this to Ember
“Create a bot that uses the Strategy A/B Test skill, and show me the plan first.”
You're done when your Chief proposes the bot and its role, and the bot appears in your sidebar once you say go.
Step 2
You're done when you see the bot’s skills and the library.
Step 3
You're done when the skill shows as on for that bot.
Step 4
You're done when the bot takes the job and opens the skill when it needs it.
Prefer to do it yourself? Choose + in the sidebar, then New Bot.Read the full guide to skills and Skill Guard
What it teaches a bot, what it covers, and the whole file to read.
The principle
You are running two Pine strategies head-to-head on the same symbol/timeframe and reporting a data-backed winner.
What it covers
Step 1: Lock the Environment
Step 2: Define What "Winner" Means
Step 3: Run Strategy A
Step 4: Restore, Then Run Strategy B
Step 5: Compare Side-By-Side
Step 6: Declare Winner + Caveats
Step 7: Report + Cleanup
Error Notes
abtest-baselineExample baseline: ES1!, 15m, Candles.
Ask the user (or use defaults) for the ranking metric. Common options:
Net Profit — simplest, ignores risk
Sharpe Ratio — risk-adjusted, preferred
Max Drawdown — lower is better (use absolute value)
Profit Factor — gross profit / gross loss
Composite — rank each metric, sum ranks, lowest wins
Default: report all four, pick winner by Sharpe, break ties by lower max drawdown.
Load A's source: paste via pine_set_source (or pine_open if saved by name)
pine_smart_compile — auto-detect type, compile, surface errors
If errors: iterate via the pine-develop loop until clean
strategy_sweep — parameter sweep with cooldown + resume. Pass the param grid (e.g., { ema_len: [20, 50, 100], atr_mult: [1.5, 2.0, 2.5] })
Collect results — strategy_sweep returns per-combination {net_profit, sharpe, max_dd, profit_factor, trades}
Example — strategy A (trend-follow):
pine_set_source(<A source>)
pine_smart_compile()
strategy_sweep({ grid: { ema_len: [50, 100, 200], atr_mult: [1.5, 2.0] } })Pick A's best row by the chosen metric. Record it.
state_restore with tag abtest-baseline — wipe A off the chart
Load B's source: pine_set_source + pine_smart_compile
strategy_sweep over B's param grid
Pick B's best row
Example — strategy B (mean-reversion):
state_restore({ tag: "abtest-baseline" })
pine_set_source(<B source>)
pine_smart_compile()
strategy_sweep({ grid: { rsi_len: [7, 14, 21], rsi_ob: [70, 75, 80] } })If strategy_sweep pauses mid-sweep due to cooldown, it will auto-resume — just wait.
Produce a metrics table:
| Metric | Strategy A (best) | Strategy B (best) | Winner |
|---|---|---|---|
| Net Profit | $8,420 | $6,110 | A |
| Sharpe | 1.42 | 1.78 | B |
| Max DD | -$2,100 | -$980 | B |
| Profit Factor | 1.65 | 1.89 | B |
| Trades | 84 | 142 | — |
Pull each row from each strategy's data_get_strategy_results for the best-params run.
Winner by default rule (Sharpe): Strategy B
Caveats:
Trade count — if B has far more trades, is it over-trading? Factor commission
Regime — was the test window mostly trending or chop? A's edge may only show in trends
Robustness — check variance across the sweep, not just the single best row. Cherry-picked best params can overfit
Show the table + winner call + caveats
capture_screenshot of the winning strategy loaded so the user has a visual
state_restore back to abtest-baseline if the user wants their original chart
Optionally state_delete the baseline snapshot once confirmed
PINE_COMPILE_ERROR — fix source, retry compile before sweeping
STRATEGY_SWEEP_COOLDOWN — expected; the tool resumes automatically
CHART_LOADING after restore — wait 1-2s and continue
SKILL.md
103 lines · 603 words
The text of SKILL.md is shown as it downloads, under the Apache-2.0 license, by Ferrox Labs. The license is the one in the file's own frontmatter.
Your new bot
Strategy A/B Test bot
Waiting in your sidebar once you say go.
No account needed · Runs on the AI plan you already pay for