Search MemoryGraph and LanceDB to surface relevant memories for user queries.
npx skills add https://github.com/ste-bah/archon --skill recall-ste-bah
Please help me install this Agent Skill. Skill: recall Source: https://github.com/ste-bah/archon/tree/main/.claude/skills/recall Command: npx skills add https://github.com/ste-bah/archon --skill recall-ste-bah
Retrieve relevant context across multiple memory stores to inform decisions without manual searching.
Ask for a topic like 'authentication timeout handling' to fetch relevant memories across all stores.
High-intent search queries and answers about installing and using this skill.
Cross-store memory search retrieves past decisions and code references by querying both MemoryGraph and LanceDB simultaneously, using keyword, fuzzy, and semantic vector matching to surface relevant context without manual searching.
Semantic vector search retrieves contextual notes by querying LanceDB to match meaning, while MemoryGraph adds keyword and fuzzy matching to capture synonyms and related concepts, together surfacing relevant memories across all stores.
Yes, fuzzy keyword matching finds memories with synonyms and related concepts by querying MemoryGraph, complementing LanceDB semantic search to ensure relevant context is surfaced across all available memory stores.
Cross-store memory search works by querying MemoryGraph for keyword and fuzzy matches alongside LanceDB for semantic vector matches, then presenting results in labeled sections to answer user queries with relevant context.
No external dependencies are required to run cross-store memory recall; the Skill operates independently to search MemoryGraph and LanceDB and surface relevant memories without additional package installations.
Yes, after retrieving memories from all stores, optional deep recall and restoration workflows are supported to restore previous context and guide current tasks with surfaced results.
caobingshengStore and recall contextual memories with frontmatter summaries and created dates.
23blocks-OSRetrieve prior discussions and decisions via memory-search.sh before starting new tasks.
winsorllcStore and retrieve semantic memories using OpenAI embeddings in a local JSON file.
hiyenwongStore and search chat histories locally with LanceDB vector embeddings and RAG retrieval.
