Skip to main content
Glama

🧠 EasyBrain

Per-project persistent memory engine, Official MCP Server & Agent Skill.
Install once globally. Keep every project's memory isolated β€” 100% locally (Zero-API-Key) or in your personal Supabase (pgvector).

License: MIT TypeScript Node.js MCP SDK


🌐 Languages / Idiomas:
English πŸ‡ΊπŸ‡Έ β€’ PortuguΓͺs (Brasil) πŸ‡§πŸ‡·


πŸ” What is EasyBrain?

EasyBrain gives your AI coding agents (Antigravity, Claude Code, Codex, Cursor, Windsurf, or custom orchestrators) persistent, isolated memory per project without bloating your context window.

When you start a new chat in the same repository, your agent calls easybrain_bootstrap to load a compact Project State Snapshot (objective, stack, decisions, current task, next steps) in ~1,000 tokens instead of re-reading the entire codebase.


Related MCP server: Sugar Memory

✨ Core Features (v1.2.1)

  • Strict Per-Project Isolation (projectId): Automatically resolves .easybrain/project.json per repository (supporting monorepos). Project A never leaks into Project B.

  • Two Processing Modes:

    • Basic Mode (Zero API Key / Zero Cost): Works out of the box with no OPENAI_API_KEY. Provides structured Project Checkpoints, instant bootstrap, lexical keyword recall, and verbatim source log tracking.

    • Enriched Mode (AI Librarian + Semantic Vectors): Add any OpenAI-compatible key (OpenAI, Groq, DeepSeek, Ollama, OpenRouter) to automatically consolidate raw logs into dense episodic summaries and vector embeddings.

  • Dual Storage Engine:

    • Local Zero-Config (STORAGE_PROVIDER=local): Stores memory per project in <project>/.easybrain/memory.json with cross-process file locking (.lock), corruption-proof schema validation, and atomic write-then-rename durability.

    • Personal Supabase (STORAGE_PROVIDER=supabase): Connects to your own Supabase (PostgreSQL + pgvector HNSW) with project_id isolation and strict lock_token atomic transactions.

  • Official MCP Server + SKILL.md: Built with @modelcontextprotocol/sdk exposing 4 lean tools (easybrain_bootstrap, easybrain_save, easybrain_recall, easybrain_get_sources) paired with a standard skills/easybrain/SKILL.md.


πŸš€ Quickstart (Zero-Setup in 60 Seconds)

git clone https://github.com/cavizu/easybrain.git
cd easybrain
npm install
npm test
npm run demo

No database account and no API key are required to run npm test or npm run demo in Basic Mode.


πŸ’» CLI Commands

# Initialize EasyBrain explicitly for a project (creates .easybrain/project.json)
npx easybrain init [--project <path>]

# Diagnose environment, active storage engine, processing mode, and project scope
npx easybrain doctor [--project <path>]

# Print ready-to-copy MCP server JSON config and Skill paths for your IDE
npx easybrain setup

πŸ› οΈ Using as an MCP Server + Agent Skill

Run the setup helper to see the exact paths for your machine:

npm run build
npx easybrain setup
  1. Install the Skill (skills/easybrain/SKILL.md) into your IDE's personal skills folder:

    • Antigravity: ~/.gemini/config/skills/easybrain/SKILL.md

    • Claude Code: ~/.claude/skills/easybrain/SKILL.md

    • Codex: ~/.agents/skills/easybrain/SKILL.md

  2. Register the MCP Server in your IDE config (claude_desktop_config.json, .cursor/mcp.json, etc.):

{
  "mcpServers": {
    "easybrain": {
      "command": "node",
      "args": ["/absolute/path/to/easybrain/dist/mcp/server.js"],
      "env": {
        "STORAGE_PROVIDER": "auto"
      }
    }
  }
}

πŸ—„οΈ Optional: Connecting Your Personal Supabase

  1. Open your Supabase SQL Editor and run sql/schema.sql.

  2. Configure .env:

STORAGE_PROVIDER=supabase
SUPABASE_URL=https://your-project.supabase.co
SUPABASE_SERVICE_ROLE_KEY=your-service-role-key
# Optional for Enriched Mode:
OPENAI_API_KEY=sk-...

🐳 Docker & REST API (Optional)

You can also run EasyBrain as a standalone REST microservice:

docker compose up -d
curl http://127.0.0.1:3000/health

The REST API exposes the same endpoints as the client: /health, /api/memory/bootstrap, /api/memory/checkpoint, /api/memory/ingest, /api/memory/retrieve, and /api/memory/logs/decompress.


πŸ“œ License

MIT License β€” see LICENSE.

Tool Schema Changelog

Recent tool additions, removals, and schema changes observed during successful MCP inspections.

Loading tool schema changelog…

    Related MCP Connectors

    Related MCP Servers

    • F
      license
      Not graded
      quality
      A
      maintenance
      Gives AI coding agents persistent, local-first memory so they can store, semantically search, and recall decisions, preferences, error patterns, and cross-project guidelines directly during a session. It also optionally queues tasks for autonomous execution against that same memory.
      -
    • F
      license
      Not graded
      quality
      A
      maintenance
      Provides AI coding agents with a local-first persistent memory layer they can read and write during sessions, storing decisions, preferences, error patterns, and research per project or globally. Enables semantic and keyword search, formatted context recall, and cross-project guideline surfacing without API keys or cloud storage.
      -
    • F
      license
      Not graded
      quality
      A
      maintenance
      Provides a local-first persistent memory layer that lets AI coding agents store, recall, and semantically search project-specific and cross-project knowledge across sessions, so context, decisions, and guidelines survive between conversations. Runs entirely on the user's machine with SQLite and no API keys, and optionally supports autonomous task-queue execution built on the same memory.
      -
    • F
      license
      Not graded
      quality
      A
      maintenance
      Gives AI coding agents a local-first persistent memory they can read and write directly during sessions, storing decisions, preferences, error patterns, and guidelines in per-project and global SQLite stores. Enables semantic and keyword retrieval of relevant context across sessions so agents start work already knowing a codebase's conventions and history.
      -