Search, recall, and store vectorized memories with file-based fallback.
npx skills add https://github.com/komluk/scaffolding --skill semantic-memory-mcp
Please help me install this Agent Skill. Skill: semantic-memory-mcp Source: https://github.com/komluk/scaffolding/tree/main/skills/semantic-memory-mcp Command: npx skills add https://github.com/komluk/scaffolding --skill semantic-memory-mcp
Agents often miss relevant past findings, decisions, and bug root-causes because semantic memory is unused or underutilized; this Skill ensures agents proactively search, recall, and store concise vectorized memories so future investigations and decisions are faster and more accurate.
Ask the agent to proactively search and recall related memories for your task and, if it discovers a verified reusable insight, store it with tags and a brief rationale.
High-intent search queries and answers about installing and using this skill.
Enable agents to proactively run semantic_search and semantic_recall operations at task start or during unexpected errors to retrieve past debugging decisions and relevant context from the semantic vector memory.
Use semantic_store write operations to save verified insights by attaching content_type, tags, agent_name attribution, and a brief rationale, ensuring future agents can effectively query and reuse the stored context.
Multi-agent workflows handle architecture decisions by applying semantic memory policies across agents, triggering semantic_search and semantic_recall to surface relevant historical context and recurring patterns before finalizing design choices.
Yes, semantic memory management supports file-based fallback when MCP services are unavailable, allowing agents to maintain read and write operations for context storage without service interruption.
You can prevent speculative noise by applying quality gates that block semantic_store operations for unverified data, ensuring only confirmed, reusable insights with clear rationale are committed to the semantic store.
Agents should use semantic_recall to retrieve formatted memory entries after semantic_search identifies potential matches, optimizing recurring pattern discovery by extracting precise context rather than broad vector matches.
aaron-he-zhuManage two-layer memory with hot cache and cold storage for SEO projects.
yanochka11Record successful technical fixes as reusable debugging recipes.
AJBcodingCoordinate multi-agent tasks and workflow orchestration in Claude Code environments.
jtucker9Retrieve past CC session details via semantic vector search with LanceDB and SQLite fallback.
