Skip to content
MCP server

Graphicle

Local code graph and project memory for coding agents, cited to exact files and lines

Runs locallyStatus and history
Pricing
Open source
Platforms
macOS, Linux, Windows
Runs
Locally
Version
0.1.18
Listing updated
3 days ago

About Graphicle

Graphicle is an open-source MCP server for developer tools. It runs locally on your machine and is started by your AI client.

Graphicle replaces repeated grepping with a query against a code graph built on your own machine. The graph is parsed with tree-sitter and no model, and it keeps Graphify's tool names such as query_graph, get_neighbors and shortest_path. The README describes the code-graph engine as a Rust port of Graphify and says the project is not affiliated with Graphify Labs.

Beside the graph it keeps project memory: Markdown, PDF, Office and HTML documents, reviewed decisions and captured Claude Code and Codex sessions. A memory is tied to the functions it concerns and goes stale when that code changes. Memories an agent proposes wait for your approval in a local browser app. Syncing happens before each answer and Git hooks refresh the graph after commits, merges and checkouts.

Everything lives in one SQLite store, with no Python, Docker or database server. By default nothing leaves the machine apart from a daily release check, and a model is needed only for written answers and proposed memories. The README reports a Kubernetes build in 3 min 20 s against 33 min 11 s for Graphify on one 4-vCPU machine.

Install

Run in a terminal

claude mcp add graphicle -- npx -y https://github.com/Graphicle-AI/graphicle/releases/download/v0.1.18/graphicle-0.1.18-aarch64-apple-darwin.mcpb

Add to claude_desktop_config.json

{
  "mcpServers": {
    "graphicle": {
      "command": "npx",
      "args": [
        "-y",
        "https://github.com/Graphicle-AI/graphicle/releases/download/v0.1.18/graphicle-0.1.18-aarch64-apple-darwin.mcpb"
      ]
    }
  }
}
Install in CursorOpens Cursor, which asks you to confirm. Or add it by hand:

Add to ~/.cursor/mcp.json

{
  "mcpServers": {
    "graphicle": {
      "command": "npx",
      "args": [
        "-y",
        "https://github.com/Graphicle-AI/graphicle/releases/download/v0.1.18/graphicle-0.1.18-aarch64-apple-darwin.mcpb"
      ]
    }
  }
}
Install in VS CodeOpens VS Code, which asks you to confirm. Or add it by hand:

Add to .vscode/mcp.json

{
  "servers": {
    "graphicle": {
      "type": "stdio",
      "command": "npx",
      "args": [
        "-y",
        "https://github.com/Graphicle-AI/graphicle/releases/download/v0.1.18/graphicle-0.1.18-aarch64-apple-darwin.mcpb"
      ]
    }
  }
}

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

{
  "mcpServers": {
    "graphicle": {
      "command": "npx",
      "args": [
        "-y",
        "https://github.com/Graphicle-AI/graphicle/releases/download/v0.1.18/graphicle-0.1.18-aarch64-apple-darwin.mcpb"
      ]
    }
  }
}

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

{
  "mcpServers": {
    "graphicle": {
      "command": "npx",
      "args": [
        "-y",
        "https://github.com/Graphicle-AI/graphicle/releases/download/v0.1.18/graphicle-0.1.18-aarch64-apple-darwin.mcpb"
      ]
    }
  }
}

What it does

Cited code graph
Files, functions, classes, calls and imports, with excerpts that name their file and lines.
Project memory
Docs, decisions and past agent sessions are searched together with the code.
Always current
Syncs changed files before answering and updates through Git hooks.
One-command setup
graphicle setup builds the graph and connects the clients it finds, backing up each file it edits.
Token-budgeted context
Answers come back as excerpts within a budget you set, with no model call needed.
Local browser app
graphicle open shows the graph, sources and memories, and is where memories are approved.
Works with many clients
Claude Code, Codex, Cursor, VS Code, Claude Desktop and other MCP clients.

Good for

  • Find where a behaviour such as request retrying is implemented
  • List what a change to a function would affect
  • Give a coding agent past decisions alongside the code
  • Review a branch with cited evidence
  • Run an agent over a very large repository without grepping it

Pricing

Open source

Free under Apache-2.0; team edition in private preview

Graphicle FAQ

Does anything leave my machine?

Not by default, apart from a daily check for a new release that can be switched off; the network is used only for models or connectors you configure.

Do I need an API key or a model?

Not for the code graph, sync, context, the MCP server, session capture or the app. A model writes answers for graphicle ask and proposes memories.

What licence is it under?

Apache-2.0, free for individuals and teams; a commercial team edition is in private preview.

How do I undo setup?

Run graphicle disconnect, which removes what setup added; each changed file is backed up first.

How big a repository can it handle?

By default up to 100,000 text files and 1 GiB of text, and both limits can be raised.

Is Graphicle free?

Yes. Graphicle is open source, so you can use and self-host it for free.

How do I add Graphicle to Claude Code?

Run this command in your terminal: claude mcp add graphicle -- npx -y https://github.com/Graphicle-AI/graphicle/releases/download/v0.1.18/graphicle-0.1.18-aarch64-apple-darwin.mcpb

What are the best alternatives to Graphicle?

The closest alternatives to Graphicle are REA, Codebase Memory, GitHub, Serena MCP: the IDE for your agent and XERJ.

Alternatives to Graphicle

  1. REA

    MCP server

    CLI and MCP server that lets coding agents reverse engineer apps and binaries

    Runs locally
    • Open source
    • 464K GitHub stars
    • Developer tools
  2. 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
  3. GitHub

    MCP server

    Official MCP server connecting AI assistants to GitHub repos, issues and workflows

    Not checked yet
    • Remote
    • Open source
    • 34.5K GitHub stars
    • Developer tools
  4. MCP toolkit giving coding agents symbol-level code retrieval and editing

    Runs locally
    • Open source
    • 31.5K GitHub stars
    • Developer tools
  5. 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
  6. Ouroboros

    MCP server

    Spec-first layer that pins requirements and verifies results of AI coding agent work

    Runs locally
    • Open source
    • 6.4K GitHub stars
    • Developer tools