Stats
Actions
Tags
From shared-memory
Retrieve semantic sharedMemory suggestions using the memory_suggest MCP tool. Get context-aware recall across project-specific and general entries without knowing exact keys.
How this skill is triggered — by the user, by Claude, or both
Slash command
/shared-memory:memory-suggestThe summary Claude sees in its skill listing — used to decide when to auto-load this skill
Use `memory_suggest` for context-based recall when you have a question or task context but don't know which memories are relevant.
Use memory_suggest for context-based recall when you have a question or task context but don't know which memories are relevant.
context: non-empty task, question, or context text (e.g., "How do I debug ripgrep performance?" or "Resume optimization strategies").tags: optional tag filter (e.g., ["learning"] or ["decision", "sharedmemory"]).limit: optional integer from 1 to 20 (default 5).When enabled (MEMORY_SUGGEST_ENABLED=true), the server uses semantic embeddings to find related entries by meaning, not just keyword match. Returns metadata only; call memory_get for full values.
Works across both project and general memories — can find learnings related to project work or project decisions related to general workflows.
Suggestions are disabled by default to avoid downloading ML models. If disabled, the tool returns an empty result without loading anything. Results are metadata-only; use memory_get for full entry values.
Success returns { ok: true, enabled, suggestions }. Disabled mode returns { ok: true, enabled: false, suggestions: [] }. Suggestions include key, summary, tags, importance, and a relevance score.
npx claudepluginhub pruthuraj/sharedmemory --plugin shared-memoryGuides creation, editing, and verification of skills for AI coding agents using test-driven development with subagent scenarios. Use when authoring or debugging skills.