Save a memory to long-term storage so it survives the conversation. Content is routed automatically to a procedure, episode, core memory, or fact, and the stored memory id is returned.
Apache 2.0
700,984 tools. Updated 2026-10-25 08:29
"Understanding Memory or Memory-Related Topics" matching MCP tools:
- Record an outcome as a reflection memory to update confidence, link related memories, and reinforce or contradict prior knowledge.MIT
- Lists all memory volumes (projects) accessible by your API key, each serving as an independent memory space for persistent memory management.MIT
- Evaluate workspace memory store health: receive a quality score, memory count, and breakdown of stale, duplicate, overlong, or never-retrieved memories. Use to diagnose memory quality problems.ISC
- Check macOS memory health with detailed stats: free, active, swap, and memory tier. Identifies pressure levels before builds or heavy tasks.Sustainable Use v1.0
- Traverse the memory graph to return connected facts around a topic, revealing surrounding context and related constraints that isolated searches miss.MIT
Matching MCP Servers
- FlicenseNot gradedqualityCmaintenanceSlipstream - Built by Vektor Memory - a persistent multi-layered memory architecture. 4-layer associative graph memory (MAGMA) with autonomous REM cycle, CLI, DXT-MCP Cloak Tools.5-
- FlicenseNot gradedqualityDmaintenanceAuto-captured, auto-recalled, path-scoped memory for AI coding agents and teams.2-
Matching MCP Connectors
- Pryzm MemoryOAuth
Encrypted long-term memory shared across your AI assistants (Claude, ChatGPT, Perplexity).
- memoryOAuth
Persistent long-term memory for AI agents: semantic search, knowledge graph, and task canvas.
- List active memory label schema to reuse existing kinds and topics, preventing near-duplicates.MIT
- Remove a stale or incorrect memory by soft-deleting it with its key; the memory is hidden from recall but remains recoverable.MIT
- Merge related episodic memories into a unified semantic memory while preserving originals as provenance. Use when multiple memories cover the same topic.MIT
- Semantic (meaning-based) memory search across all six memory layers via Atlas Vector Search autoEmbed. Finds conceptually related memories even without keyword overlap. Complements keyword-based recall/query_layer.MIT
- Store or update a JSON value for a memory key with optional TTL, importance, source, and confidence score. Persists AI agent memory across sessions and changes.MIT
- Reinforce or deny a stored memory pattern by confirming when user validates or denying when they contradict it for accurate AI memory persistence.MIT
- Merge a reviewed draft into canonical memory, optionally marking an older memory as superseded.MIT
- Delete a specific memory by providing its unique UUID, allowing removal of outdated or incorrect information from the shared memory store.MITDestructive
- Reduce memory bloat by merging near-duplicate memories and summarizing related clusters into concise long-term memories. Compress episodic experiences into semantic knowledge.MIT
- Delete a specific memory by providing its UUID. Removes unwanted or outdated records from the memory store.MITDestructive
- Retrieve all incoming and outgoing links for a memory, optionally filtered by relation type, to view connections in the memory graph.MIT
- Retrieve memories related to an anchor memory, categorized as forward, backward, or lateral relationships based on vector similarity.MIT
- Changes the database's name, description or memory. (scope: databases:write)MIT
- Detect memory-poisoning risks in agent code: unsanitized untrusted input in persistent memory and memory retrievals used as trusted context.MIT
- Fetch a ClawMem memory using an issue number or memory reference string. Retrieve specific stored data for your AI agent.MIT