| name | memory-extraction |
| description | Automatic memory extraction skill — triggered by file operations to extract durable learnings from session context. Runs in the background via the memory-extractor agent.
|
| user-invocable | false |
| metadata | {"version":"2.0","effort":"low","auto-invocable":true,"category":"auto","compatible-claude-code":{"when_to_use":"Auto-triggered after significant session milestones","user-invocable":false,"context":"fork","allowed-tools":["Read","Write","Edit","Grep","Glob"]}} |
Auto Memory Extraction
This skill runs automatically to extract durable memories from session context.
It is NOT user-invocable — it is triggered by the memory system.
How It Works
- The
@memory-extractor subagent reviews session context
- It extracts: decisions, corrections, preferences, connections, tools
- It writes to
memory/ directory topic files
- It updates
memory/MEMORY.md index
Manual Trigger
Use /memory-sync to manually trigger memory extraction from the current session.
Files
memory/MEMORY.md — topic indexmemory/connections.md — hosts, ports, endpointsmemory/decisions.md — architecture choicesmemory/lessons.md — corrections and gotchasmemory/preferences.md — workflow preferencesmemory/tools.md — discovered tools and utilities
Rules
- Never store secrets (API keys, passwords, tokens)
- Keep entries concise — one line per memory
- Read before writing to avoid duplicates
- Delete stale info when noticed
