mnemostack
Related Servers
Alternatives to mnemostack
No user-submitted related servers found.
Related Servers
- AlicenseNot gradedqualityCmaintenanceMCP server that enables persistent, hybrid, local memory for LLM agents, with vector + BM25 search, knowledge graph, and policy-driven retention, providing token-budgeted context injection for AI assistants.MIT

Kagura Memory Cloudofficial
AlicenseNot gradedqualityAmaintenanceProvides AI agents and teams with self-hosted adaptive memory, combining hybrid search and a self-improving neural memory graph to enable precise recall and serendipitous discovery across MCP-compatible clients.12Apache 2.0- AlicenseAqualityBmaintenanceLocal-first agent memory & knowledge service: hybrid retrieval (vector + BM25/RRF) + memory governance (decay, dedup, freshness), REST + MCP dual protocol, fully offline without LLM. Python.82Apache 2.0
- AlicenseNot gradedqualityDmaintenancePersistent, semantically-searchable memory for AI agents using local PostgreSQL, pgvector, and Ollama embeddings, exposed via MCP with hybrid retrieval, knowledge graph, and auto-recall hook.3 npmMIT
- AlicenseNot gradedqualityBmaintenanceEnables agents to maintain multi-tier memory across working scratchpad, dialogue buffer, and episodic stores with recency decay. Supports hybrid BM25/vector retrieval, rank fusion, graph expansion, and context reordering for optimized MCP-based memory and RAG workflows.7MIT
- AlicenseAqualityAmaintenancePersistent memory MCP server for AI coding agents (Claude Code, Codex, Gemini CLI). Hybrid retrieval (vector + BM25), cross-encoder reranking, knowledge graph, session checkpoint/resume, and multi-scope isolation. Local-first with LanceDB.3085 npm15MIT
TDQS
Scored across 7 tools
Each tool has a clearly distinct purpose: health checks, raw search, synthesized answer, citation resolution, invalidation, memory write, and feedback. The descriptions explicitly differentiate overlapping-sounding pairs like search vs. answer and invalidate vs. remember. No two tools appear to do the same thing.
All tools use a consistent mnemostack_ prefix and snake_case, making them easy to predict. However, the action words mix verbs (search, resolve, invalidate, remember) with nouns (health, feedback), a minor deviation from a pure verb_noun pattern.
Seven tools is well-scoped for a memory backend covering health, read/write, feedback, and citation verification. Each tool earns its place without redundancy or thinness. The count falls comfortably in the ideal 3–15 range.
The set covers core lifecycle operations: store, search, answer, invalidate, resolve citations, and record feedback. Minor gaps include no direct update tool (must invalidate and re-remember) and the referenced mnemostack_graph_add_triple is not exposed, but these are workable or intentionally out of scope.