recall

Installation
SKILL.md

/recall

Query persistent memory shared between all agents (claude-code, opencode, clawdbot) using hybrid search: 70% semantic vector similarity + 30% BM25 keyword matching.

When You Need This (And When You Don't)

At session start, Signet automatically injects relevant memories into your context — scored by importance, recency, and relevance to the current conversation. MEMORY.md is also regenerated periodically from the full database using decay-weighted scoring. For most interactions, the right memories are already present without any manual searching.

Use /recall when you need to:

  • Search for something specific that isn't in current context
  • Look up old decisions or past conversations
  • Debug whether a memory was actually captured
  • Find memories by type, tag, or date range
Related skills

More from signet-ai/signetai

Installs
1
GitHub Stars
140
First Seen
Apr 9, 2026