ai-notelisted

# Note ## Purpose Knowledge management for technical discoveries. Captures debugging insights, non-obvious behaviors, and integration gotchas that would cost 30+ minutes to re-discover. Notes persist across sessions and are searchable. ## Trigger - Command: `/ai-note <slug>` (create/update) or `/ai-note find [query]` (search) - Context: discovery during debugging, research, or investigation that should be preserved. ## When to Use - Discovery cost exceeds 30 minutes of investigation - Non-obvious behavior that contradicts documentation or expectations - Debugging insight that required multiple attempts to isolate - Integration gotcha between tools, libraries, or services - Workaround for a known issue with no upstream fix ## When NOT to Use - **Architecture decisions** -- use `decision-store.json` via `/ai-governance` - **Incident analysis** -- use `/ai-postmortem` - **Customer issues** -- use `/ai-support` ## Workflow ### Mode: find 1. **Search notes** -- scan `.ai-engineering/notes/*.md` for files matching the query in filename, title, or content. 2. **Rank results** -- sort by relevance (title match > content match > date). 3. **Present** -- list matching notes with title, date, and first-line summary. ### Mode: create/update (by slug) 1. **Check existing** -- look for `.ai-engineering/notes/{slug}.md`. If found, load for update. 2. **Gather context** -- from the current session, extract: - What problem was being solved - What was tried and failed - W