Skip to content
MCP server

heropen — AI Agent Memory

Local-first long-term memory for AI agents, stored on your own machine

Runs locallyStatus and history
Install
Pricing
Free
Platforms
CLI, Windows, macOS
Runs
Locally
Version
2.0.8
Listing updated
3 days ago

About heropen — AI Agent Memory

heropen — AI Agent Memory is a free MCP server for knowledge and memory. It runs locally on your machine and is started by your AI client.

heropen keeps an agent's memories in SQLite files under ~/.heropen/, one file per agent, with nothing uploaded and no telemetry. It installs with pip install heropen, needs Python 3.10 or later, and asks for no Docker, Postgres or API key. On first launch it tries to register itself with known clients such as Claude Code, Cursor and Windsurf, while setups like Hermes need the heropen-mcp command wired in by hand.

Search is full-text out of the box, and local semantic search is an optional extra that downloads an embedding model of roughly 95 MB. If embeddings are unavailable, search falls back to full-text and storage keeps working. Other features the site lists include a browser dashboard for viewing and managing memories, session checkpoint protection when a conversation is compressed, a time-awareness note at the start of each chat, and a Windows tray badge. The free tier gives eight fully private agents plus an optional shared domain where agents exchange knowledge.

The free version has no time limit. Plus and enterprise tiers are described as coming, with shared memory domains, private deployment and permissions. The website is written in Chinese, and a JSON export is available through heropen export.

Install

Run in a terminal

claude mcp add heropen -- uvx heropen

Add to claude_desktop_config.json

{
  "mcpServers": {
    "heropen": {
      "command": "uvx",
      "args": [
        "heropen"
      ]
    }
  }
}
Install in CursorOpens Cursor, which asks you to confirm. Or add it by hand:

Add to ~/.cursor/mcp.json

{
  "mcpServers": {
    "heropen": {
      "command": "uvx",
      "args": [
        "heropen"
      ]
    }
  }
}
Install in VS CodeOpens VS Code, which asks you to confirm. Or add it by hand:

Add to .vscode/mcp.json

{
  "servers": {
    "heropen": {
      "type": "stdio",
      "command": "uvx",
      "args": [
        "heropen"
      ]
    }
  }
}

Add to ~/.codeium/windsurf/mcp_config.json

{
  "mcpServers": {
    "heropen": {
      "command": "uvx",
      "args": [
        "heropen"
      ]
    }
  }
}

Add to your client's MCP settings (most use this shape)

{
  "mcpServers": {
    "heropen": {
      "command": "uvx",
      "args": [
        "heropen"
      ]
    }
  }
}

What it does

Local storage
One SQLite file per agent under your home directory, never uploaded.
Full-text and semantic search
Full-text by default, local embeddings as an optional extra.
Private and shared domains
Eight private agents on the free tier plus an optional shared domain.
Session checkpoints
Context survives when a conversation is compressed.
Time awareness
Each session opens with local time and the gap since the last chat.
Browser dashboard
View, search and manage memories and agents from a local panel.

Good for

  • Give a coding agent memory of project conventions across sessions
  • Keep personal preferences available in every chat
  • Share notes between two agents through a shared domain
  • Back up memories as JSON and move them to another machine

Pricing

Free

Free with no time limit; Plus and enterprise tiers announced as coming

Prices change. Check heropen.net before you buy.

heropen — AI Agent Memory FAQ

Does any data leave my machine?

The README says memories are never uploaded or synced, with no telemetry or cloud sync.

Do I need an API key?

No. Installation is a single pip command with no key, Docker or database.

Does it work without embeddings?

Yes. Search falls back to full-text and storage and retrieval keep working.

What is free?

The free version has no time limit and includes basic memory, the MCP server, local storage and the dashboard.

Is heropen — AI Agent Memory free?

Yes. heropen — AI Agent Memory is free to use.

How do I add heropen — AI Agent Memory to Claude Code?

Run this command in your terminal: claude mcp add heropen -- uvx heropen

What are the best alternatives to heropen — AI Agent Memory?

The closest alternatives to heropen — AI Agent Memory are Codebase Memory, XERJ, Claude Code Ultimate Guide, Nautilus Compass Memory and note1.

Alternatives to heropen — AI Agent Memory

  1. Codebase Memory

    MCP server

    Local MCP server that builds a code knowledge graph for AI coding agents

    Runs locally
    • Open source
    • 48.7K GitHub stars
    • Knowledge and memory
  2. XERJ

    MCP server

    Local search engine for AI agents with cited answers over your code and docs

    Runs locally
    • Open source
    • 19.2K GitHub stars
    • Knowledge and memory
  3. MCP server for searching the Claude Code Ultimate Guide from any MCP client

    Runs locally
    • Open source
    • 6.5K GitHub stars
    • AI models and LLM tooling
  4. Local-first long-term memory layer for AI agents, served over MCP

    Not checked yet
    • 5 tools
    • Remote
    • Open source
    • 1.3K GitHub stars
    • Knowledge and memory
  5. note1

    MCP server

    MCP server giving AI assistants search and summaries across your note1 meetings

    • 4 tools
    • Remote
    • Free trial
    • 1 GitHub star
    • Productivity