Skip to main content
Glama
AyubMoh1

memory-mcp

by AyubMoh1

memory-mcp

Persistent memory MCP server for Claude Code. Store, search, and retrieve memories across sessions using hybrid vector + keyword search powered by SQLite, FTS5, and sqlite-vec.

Features

  • Hybrid search — 70% vector similarity + 30% BM25 keyword matching

  • Local embeddings — Ollama (required, local, free)

  • Persistent storage — SQLite database survives across sessions

  • Token-aware context — Retrieve memories fitted within a token budget

  • File watching — Auto-index markdown and code files on change

  • LRU embedding cache — 10k entries, SHA256-keyed dedup

Related MCP server: recall-mcp

Setup

git clone https://github.com/AyubMoh1/memory-mcp.git
cd memory-mcp
npm install
npm run build

Register with Claude Code

Add to ~/.claude/config.json:

{
  "mcpServers": {
    "memory": {
      "command": "node",
      "args": ["/path/to/memory-mcp/dist/index.js"]
    }
  }
}

Tools

Tool

Description

memory_store

Store a memory with category, tags, and importance

memory_search

Hybrid vector + keyword search

memory_list

List recent memories with filters

memory_delete

Delete a memory by ID

memory_get_context

Token-budgeted context retrieval for a topic

memory_get_stats

Storage statistics

Resources

Resource

URI

Description

Stats

memory://stats

JSON storage statistics

Recent

memory://recent

Last 20 memories

Configuration

All configuration is via environment variables:

Variable

Default

Description

MEMORY_DB_PATH

~/.memory-mcp/memory.db

Database file location

MEMORY_WATCH_PATHS

(empty)

Comma-separated file paths to auto-index

OLLAMA_URL

http://127.0.0.1:11434

Ollama server URL

DEBUG

Enable debug logging

Embeddings

Ollama is required. The server will fail to start without a running Ollama instance and an embedding model installed.

ollama serve                    # Start Ollama
ollama pull nomic-embed-text    # Install embedding model

Architecture

src/
├── index.ts              # Entry point: MCP server + stdio transport
├── storage/
│   ├── types.ts          # MemoryChunk, SearchResult, StorageBackend
│   └── database.ts       # SQLite + FTS5 + sqlite-vec
├── embeddings/
│   ├── providers.ts      # Ollama embedding provider
│   ├── cache.ts          # LRU embedding cache
│   └── detect.ts         # Detect Ollama + embedding model
├── sync/
│   ├── file-watcher.ts   # chokidar with debouncing
│   └── chunker.ts        # Split files by headers/paragraphs
├── tools/                # MCP tool handlers
├── resources/            # MCP resource handlers
└── utils/                # Logger, ID gen, token estimation

License

MIT

A
license - permissive license
-
quality - not tested
D
maintenance

Maintenance

Maintainers
Response time
Release cycle
Releases (12mo)
Commit activity

Resources

Unclaimed servers have limited discoverability.

Looking for Admin?

If you are the server author, to access and configure the admin panel.

Related MCP Servers

View all related MCP servers

Related MCP Connectors

  • Hosted MCP memory: save sessions/decisions once, search from Claude, Cursor, ChatGPT. EU-hosted FTS.

  • Cloud-hosted MCP server for durable AI memory

  • One memory, every AI: Claude, ChatGPT, Perplexity, Gemini, Cursor, OpenClaw, Hermes, any MCP client.

View all MCP Connectors

Latest Blog Posts

MCP directory API

We provide all the information about MCP servers via our MCP API.

curl -X GET 'https://glama.ai/api/mcp/v1/servers/AyubMoh1/memory-mcp'

If you have feedback or need assistance with the MCP directory API, please join our Discord server