orient
Load project context now — read the memory index + recent state before continuing.
Append a one-line entry to the Mistakes Log (symptom · correction · codified).
How it fires
How this command gets triggered: by you, by Claude, or both.
/log-mistakeContext preview
What this command does when you run it.
Append a one-line entry to the Mistakes Log (symptom · correction · codified).
description: Append a one-line entry to the Mistakes Log (symptom · correction · codified). argument-hint: "[what went wrong]"
Append a Mistakes Log entry (to `CLAUDE.md`'s Mistakes Log section, or your mistakes-log file) for:
**$ARGUMENTS**
Use this format, with today's date (UTC):
`YYYY-MM-DD — <symptom>. Correction: <how to avoid it next time>. Codified: <test / hook / rule, or "not yet">`
Be specific in the correction. Then ask the real question: **can this be made mechanical?** If a hook or a test could stop this class of mistake from recurring, say so in Codified and offer to build it — that's the difference between writing it down and making it un-repeatable.
Stop fixing the same AI mistake twice. A Lean quality system for AI-assisted work — operators turn recurring defects into countermeasures. Built by a manufacturing operator, for people who actually ship.
Load project context now — read the memory index + recent state before continuing.
Close the session — save state to memory, update CONTEXT, commit, surface next moves.