开始使用
Mistake Journal Guard — Pre-Trade Warnings

Mistake Journal Guard — Pre-Trade Warnings

An AI that predicts the market is everywhere. This is the opposite: a guard that remembers YOUR past trading mistakes and warns you before you repeat one. Keep a small Markdown pattern journal (FOMO chases, thesis-less averaging down, "bubble" misses); before any trade decision the agent scans it and shows your own words back — how it ended last time, and the rules you set for yourself. Runs 100% locally, no API keys, Korean & English. Self-review tool only — not investment advice.
#财务#分析#生产力
评分
需要更多评价
订单数
0
如何使用
下载

Mistake Journal Guard

Markets don't repeat, but a trader's mistakes do. This skill turns your own mistake journal into a pre-trade tripwire: before helping with any trade decision, the agent scans your journal — and if your current reasoning rhymes with a recorded pattern, it stops and shows you your own words: how it ended last time, and the rules you wrote for yourself afterward.

Then it helps you anyway. It's a mirror, not a lock.

Disclaimer: This tool mirrors your own recorded decision patterns for general self-review and educational purposes only. It is not financial or investment advice and does not constitute a recommendation to buy, sell, or hold any asset. This agent does not constitute professional financial advice - consult a licensed financial advisor before making financial decisions. Market conditions are uncertain, and past patterns do not guarantee future results.

How it works

Your journal is a small Markdown file - a pattern dictionary, not a diary. One section per recurring mistake, each with:

  • Trigger - the situation where the mistake fires
  • Rationalization - the exact phrases running through your head at that moment
  • Outcome - how it usually ends
  • Warning signs - cues the agent can detect in your words
  • Defenses / rules - what you told yourself to do next time
  • Cases - dated one-line retrospectives

When you mention buying, selling, adding, or cutting a position, the agent parses the journal, runs a keyword prefilter, then semantically compares your stated reasoning against your recorded rationalizations. On a match it warns first - quoting your journal verbatim - then continues with whatever you asked. It never blocks, never moralizes, and a clean result is never an endorsement of the trade.

What's included

  • SKILL.md - the full agent protocol (scan → match → warn → help)
  • scripts/guard.py - stdlib-only Python CLI: init, list, show, match, add-case, stats, JSON output for agents
  • Journal templates in English and Korean (오답노트), pre-loaded with three classic starter patterns: FOMO chasing, thesis-less averaging down, and dismissing a structural trend as "a bubble"

Recording

After a loss or a "did it again" moment, add a dated case line to an existing pattern (add-case) or author a new pattern with the agent's help. stats ranks patterns by case count - the top one is your most expensive habit, useful for weekly or quarterly reviews.

Privacy & framing

  • Runs 100% locally: no network calls, no API keys, no telemetry. Your journal never leaves your disk.
  • Works with any Markdown setup - Obsidian vaults included.
  • Every analysis output ends with a fixed financial disclaimer. The guard judges your decision process against your own records - never the trade itself, in either direction.