joshuaswarren / remnic-search

When to use

Use when remnic-recall did not surface something the user insists exists, or when the task genuinely needs exhaustive coverage rather than a ranked semantic summary.

Triggers:

• "Search memories for …"
• "Deep search on …"
• /remnic:search <query> slash command invocation.
• Follow-up after remnic-recall returned nothing useful.

Inputs

• query (required) — literal phrase or keyword string; this is full-text, not semantic.
• Optional: date range, category filter, entity constraint.

Procedure

1. Pick the most literal phrase the user expects to match.
2. Call remnic_lcm_search with that phrase.
3. Group results by date or category when the tool returns enough metadata.
4. Present the top 5–10 matches with dates and short excerpts.
5. If still nothing, say so plainly and offer remnic-remember for capturing the content going forward.

Efficiency plan

• Use the most specific phrase available.
• One targeted search beats several broad ones.
• Do not re-run a deep search for the same topic in the same turn after recall already covered it.

Pitfalls and fixes

• Pitfall: Using remnic_lcm_search as the default. Fix: Start with remnic_recall; escalate only when recall fails.
• Pitfall: Pasting huge result dumps. Fix: Show top matches with excerpts.
• Pitfall: Over-broad queries. Fix: Require at least one distinctive keyword.

Verification checklist

• [ ] remnic_lcm_search was called only after recall fell short or exhaustive matching was required.
• [ ] Results were grouped by relevance or date when possible.
• [ ] Output shows excerpts, not raw payloads.
• [ ] Canonical remnic_lcm_search was used over legacy engram_lcm_search.

Tool names: canonical name is remnic_lcm_search. The legacy engram_lcm_search alias remains accepted during v1.x.