mem-search

3-layer workflow. Do not dump full observations until IDs are filtered.

1. search(query) — index with IDs
2. timeline(anchor=ID) — nearby context
3. get_observations(ids=[...]) — full details for those IDs only

Stamp platformSource as cursor or grok-bot on writes for this host. When reading, do not drop the other host unless asked.

If MCP is missing, run the install skill first (npx claude-mem install --ide <host>).