Skip to main content
Glama
mnemostack
by udjin-labs

Related Servers

Alternatives to mnemostack

No user-submitted related servers found.

    Related Servers

    • A
      license
      Not graded
      quality
      D
      maintenance
      MCP 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
    • A
      license
      Not graded
      quality
      B
      maintenance
      Provides 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.
      13
      Apache 2.0
    • A
      license
      A
      quality
      B
      maintenance
      Local-first agent memory & knowledge service (v2.4.4): hybrid retrieval (vector + BM25) + offline CSV/Excel row-level Q&A without LLM; REST + MCP dual protocol (13 tools incl. list/delete documents); memory governance; Python.
      13
      62 PyPI
      5
      Apache 2.0
    • A
      license
      Not graded
      quality
      D
      maintenance
      Persistent, 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.
      MIT
    • A
      license
      Not graded
      quality
      B
      maintenance
      Enables 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.
      7
      MIT
    • A
      license
      A
      quality
      B
      maintenance
      Persistent 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.
      30
      26 npm
      15
      MIT

    TDQS

    A3.9/5.0

    Scored across 7 tools

    Disambiguation4/5

    Each tool maps to a distinct operation: health, raw search, synthesized answer, store, retract, feedback, and citation resolve. The only close pair is search vs. answer, but the descriptions explicitly contrast 'raw memory matches' against 'a synthesized answer,' so an agent can still choose correctly.

    Naming Consistency4/5

    All tools share a uniform mnemostack_ prefix with snake_case single-word suffixes. Minor deviation: most suffixes are verbs (search, answer, remember, invalidate, resolve) but health and feedback are nouns, a small inconsistency in an otherwise predictable scheme.

    Tool Count5/5

    Seven tools is well within the ideal 3-15 range for a memory backend. Each tool earns its place with no redundant surface: read paths (health, search, answer, resolve), write paths (remember, invalidate), and learning (feedback).

    Completeness4/5

    The lifecycle is largely covered: store (remember), retrieve (search/answer), retract (invalidate), verify (resolve), learn (feedback), and health. Gaps remain around explicit update/edit of a memory, hard delete, and listing/enumerating stored memories or collections, but these are workable minor omissions.

    Maintenance

    ActivityActive
    ResponsivenessUsually handled