trade-journal
Skill2 starsChanged 5 months agoMITView on GitHub
Capture the psychological side of a trade - emotion, conviction source, confidence, and plan adherence - at entry and exit, then surface which felt-states actually lose money. Use when the user says "journal this trade", "log my entry", "I just bought/sold X", "reconcile my trade", "trade journal", "why do I keep losing", "what's my real psychological pattern", or wants emotional/behavioral review of their trading. Pairs with pre-trade-check (entry) and weekly-mirror (review).
10 other projects use a version of this.
Adopted by
Most recently taken up by
- PatrickSUDO/fadacai-portfoliotheir own version, first seen Nov 13, 2026
- nimadorostkar/Claude-Skills-collectiontheir own version, first seen Oct 13, 2026
- bocodes1/claude-skillstheir own version, first seen Oct 11, 2026
- IngAs8/ClaudeSkilltheir own version, first seen Sep 28, 2026
- Serennity007/claude-trading-skills-67their own version, first seen Aug 25, 2026
- NT-NinjaTrader/mcp-skillstheir own version, first seen Jul 25, 2026
Other versions10 projects use a version of this.
Word for word · 0
None found.
Edited · 0
None found.
Wrote their own version · 10
- agiprolabs/claude-trading-skills★ 366same name
- PatrickSUDO/fadacai-portfolio★ 145same name
- nimadorostkar/Claude-Skills-collection★ 37same name
- ZICXR/A-Stock-Skills★ 26same name
- jackson-video-resources/skills★ 17same name
- NT-NinjaTrader/mcp-skills★ 8same name
- bocodes1/claude-skills★ 0same name
- IngAs8/ClaudeSkill★ 0same name
- Serennity007/claude-trading-skills-67★ 0same name
- Yahmez3/Claude-Skills★ 0same 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 repository29 other files it gives its agents.
- AGENTS.mdAGENTS.md
- CLAUDE.mdCLAUDE.md
- adversarial-researchSkill
- auto-rebalanceSkill
- cash-deploymentSkill
- daily-briefingSkill
- dividend-strategySkill
- earnings-analyzerSkill
- earnings-postmortemSkill
- employer-stockSkill
- health-checkSkill
- macro-impactSkill
- momentum-scannerSkill
- optimize-allocationSkill
- pead-trackerSkill
- perf-optimizerSkill
- portfolio-healthSkill
- portfolio-reviewSkill
- position-reviewSkill
- pre-trade-checkSkill
- profile-setupSkill
- risk-assessmentSkill
- sector-rotationSkill
- stock-analysisSkill
- stock-compareSkill
- tax-loss-reviewSkill
- top-trades-todaySkill
- trading-deskSkill
- weekly-mirrorSkill
The file
We split this skill into the rules it gives an agent
5 rules. 0 are common, 3 are distinctive, and 2 are very specific to this project.
0
Common
Most of the 9 edited versions we read keep these.
3
Distinctive
This author's own choices. Worth a look.
2
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Very specific: line 46
Step 3 - Mirror back one line: If emotion ∈ {fomo, revenge, fear} OR conviction_source ∈ {tip, social, gut} OR …
- Very specific: line 66
If it returns reconciled=false, there was no open entry - tell the user the entry was never journaled.
- Distinctive: line 104
Never invent felt-state. If the user did not state an emotion, ask.
- Distinctive: line 106
Keep entry/exit interactions to <120 words.
- Distinctive: line 107
No PII. Journal stores ticker + felt-state only.
“Common” is measured: kept by at least 6 in 10 of the 9 edited versions of this file we could read. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
Trade Journal (Psychological Layer)
Goal
Most retail losses are emotional, not analytical. decision_memory records the facts of a trade (price, conviction label, thesis) and shadow_account derives biases from price/date. Neither captures what the trader felt at the moment of entry. This skill does - emotion, conviction source, self-rated confidence, plan adherence - then get_journal_insights cross-tabs felt-state against realized outcomes to answer the one question that beats technical analysis: what emotional state precedes my losing trades?
This is a capture + mirror skill, not a recommender. It does not approve or reject trades (that's pre-trade-check).
When to invoke
- ENTRY: user just placed (or is about to place) a trade and wants it journaled. Natural follow-on from pre-trade-check.
- EXIT: user closed a position - reconcile the open journal entry.
- REVIEW: "why do I keep losing", "what's my pattern", "trade journal insights".
How to run
Step 0 - Detect phase (REQUIRED): entry, exit, or review. If ambiguous, ask.
ENTRY phase
Step 1 - Capture felt-state (ask only what the user has not already said):
- Ticker?
- Emotion right now? → one of:
calm | fomo | fear | revenge | conviction | bored | uncertain - Where did the conviction come from? →
thesis | chart | tip | social | news | gut - Confidence 1-5? (1 = a flyer, 5 = highest conviction)
- What is the plan? (entry / stop / exit stated BEFORE the trade - one sentence)
- Did pre-trade-check pass? (yes/no - if they skipped it, note that, it is itself a signal)
Step 2 - Persist:
Call mcp__aifolimizer__log_trade_journal with:
ticker=<TICKER>
emotion=<emotion>
conviction_source=<source>
confidence_1to5=<1-5>
plan_intended=<one-sentence plan>
felt_note=<free text, optional>
pre_trade_check_passed=<true/false>Step 3 - Mirror back one line. If emotion ∈ {fomo, revenge, fear} OR conviction_source ∈ {tip, social, gut} OR pre_trade_check_passed=false, flag it plainly: "Logged. Note: this is a {emotion}/{source} entry - historically your weakest setup (check insights)." Do not moralize beyond one line.
EXIT phase
Step 1 - Reconcile (ask):
- Ticker?
- Did you follow the plan you wrote at entry? (yes/no - be honest)
- Emotion at exit? (same enum)
- Was the outcome a surprise? →
expected | mild_surprise | shock - One-line lesson?
Step 2 - Persist:
Call mcp__aifolimizer__log_trade_journal_exit with:
ticker=<TICKER>
plan_followed=<true/false>
exit_emotion=<emotion>
outcome_surprise=<level>
lesson=<one line>If it returns reconciled=false, there was no open entry - tell the user the entry was never journaled; offer to log a retroactive entry.
REVIEW phase
Step 1 - Mark outcomes current, then pull insights (in order):
mcp__aifolimizer__resolve_trade_outcomes(marks decisions to market so wins/losses are fresh)mcp__aifolimizer__get_journal_insights
Step 2 - Present the mirror (use the output template). Lead with the single worst felt-state by win-rate. State the confidence-calibration verdict bluntly: if avg_confidence_losses >= avg_confidence_wins, the user's confidence is inverted - they feel most sure exactly when they are most wrong.
Output template (REVIEW)
TRADE JOURNAL - PSYCHOLOGICAL MIRROR
Scored trades: <scored_entries> of <total_entries> (<open_unreconciled> not yet closed)
WIN-RATE BY ENTRY EMOTION
conviction <n> <win_rate_pct>%
calm <n> <win_rate_pct>%
fomo <n> <win_rate_pct>% <-- worst
...
WIN-RATE BY CONVICTION SOURCE
thesis <n> <win_rate_pct>%
social <n> <win_rate_pct>% <-- worst
...
CONFIDENCE CALIBRATION
avg confidence on WINS: <avg_confidence_wins>
avg confidence on LOSSES: <avg_confidence_losses>
verdict: <calibrated | INVERTED - you are most sure when most wrong>
THE PATTERN: <one sentence naming the emotion+source that loses money>Rules
- Capture only - never approve/reject a trade (route to pre-trade-check for that).
- Never invent felt-state. If the user did not state an emotion, ask; do not guess it from price action.
- Enums are fixed - map the user's words to the nearest enum value; if none fits, ask.
- Keep entry/exit interactions to <120 words. The review is the only long output.
- No PII. Journal stores ticker + felt-state only; no balances, no account IDs.
Gotchas
get_journal_insightscan only score entries whose ticker has a resolved decision (target_hit/stop_hit). Ifscored_entriesis 0 buttotal_entries> 0, tell the user outcomes aren't resolved yet - runresolve_trade_outcomesor wait for trades to hit target/stop.- Insights need a decision logged too (via pre-trade-check / log_trade_decision). A journal entry with no matching decision is captured but unscoreable. Encourage logging both.
- Small-sample caveat: with <5 scored trades per bucket, win-rates are noise. Say so.
- This skill feeds
weekly-mirror(lessons) - do not duplicate the performance-math output; defer P&L/R-multiple to weekly-mirror.
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.

