Skip to content
MCP server

SyncytiumMD

Keep AI coding rules once and generate each tool's instruction file, with agent handoff

Runs locallyStatus and history
Install
Pricing
Open source
Platforms
CLI, macOS, Linux
Runs
Locally
Version
0.3.2
Listing updated
2 days ago

About SyncytiumMD

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

Different assistants want different instruction files, so teams end up with drifting copies of the same rules. SyncytiumMD keeps one canonical .syncytium/ folder in git and transpiles it into files such as CLAUDE.md, AGENTS.md, GEMINI.md, .cursorrules, .clinerules, .windsurfrules and Copilot instructions. Each generated file carries a banner, and syncytium diff --check fails CI when any has drifted. syncytium clean removes only banner-tagged files.

It also coordinates agents. A lease-based workspace lock with heartbeat, a handoff note saying who has the baton and what is next, and a 100-entry audit trail stop agents overwriting each other. Rules and decisions records can be searched, added and edited from the CLI, and a local 3D graph shows the knowledge structure, bound to 127.0.0.1.

The MCP server exposes 23 tools. Deletions default to dry-run and need explicit confirmation, items are archived to a trash folder so they can be restored, and per-tool rate limits are enforced server-side.

Install

Run in a terminal

claude mcp add syncytiummd -- npx -y syncytium-md

Add to claude_desktop_config.json

{
  "mcpServers": {
    "syncytiummd": {
      "command": "npx",
      "args": [
        "-y",
        "syncytium-md"
      ]
    }
  }
}
Install in CursorOpens Cursor, which asks you to confirm. Or add it by hand:

Add to ~/.cursor/mcp.json

{
  "mcpServers": {
    "syncytiummd": {
      "command": "npx",
      "args": [
        "-y",
        "syncytium-md"
      ]
    }
  }
}
Install in VS CodeOpens VS Code, which asks you to confirm. Or add it by hand:

Add to .vscode/mcp.json

{
  "servers": {
    "syncytiummd": {
      "type": "stdio",
      "command": "npx",
      "args": [
        "-y",
        "syncytium-md"
      ]
    }
  }
}

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

{
  "mcpServers": {
    "syncytiummd": {
      "command": "npx",
      "args": [
        "-y",
        "syncytium-md"
      ]
    }
  }
}

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

{
  "mcpServers": {
    "syncytiummd": {
      "command": "npx",
      "args": [
        "-y",
        "syncytium-md"
      ]
    }
  }
}

Tools this server provides (17)

Taken from the project documentation; the running server may differ.

syncytium_get_context
Returns the project context.
syncytium_handoff
Records or reads an agent handoff.
syncytium_create_rule
Adds a canonical rule.
syncytium_update_rule
Edits a rule.
syncytium_remove_rule
Removes a rule, dry-run by default.
syncytium_record_decision
Records an architecture decision.
syncytium_sync
Regenerates bridge files.
syncytium_diff
Shows drift.
syncytium_doctor
Runs read-only health checks.
syncytium_validate
Validates config and rules.
syncytium_validate_fix
Fixes validation problems.
syncytium_lock_acquire
Acquires the workspace lock.
syncytium_lock_release
Releases the workspace lock.
syncytium_lock_status
Shows lock status.
syncytium_get_graph
Returns the knowledge graph.
syncytium_list_trash
Lists archived deletions.
syncytium_restore
Restores an archived item.

What it does

Single source of rules
Author once in `.syncytium/` and sync to 10 built-in adapters.
Drift detection
`diff --check` exits non-zero in CI when generated files differ.
Agent lock and handoff
Leases, heartbeats and a handoff baton with an audit trail.
Safe MCP tools
Destructive tools default to dry-run and need confirmation.
Reverse import
Migrate existing CLAUDE.md or .cursorrules into the canonical folder.
Local knowledge graph
A 3D graph and document reader served on loopback only.

Good for

  • Maintain one ruleset for Cursor, Claude Code and Copilot
  • Fail CI when generated instruction files drift
  • Pass work between agents with a recorded handoff
  • Record architecture decisions the agents can read

Pricing

Open source

SyncytiumMD FAQ

Which tools does it generate files for?

Cursor, Claude Code, Copilot, Cline/Roo, Antigravity, Windsurf, Trae, OpenCode, AGENTS.md and GEMINI.md.

What does it require?

Node.js 20 or later; npx is enough.

Can the assistant delete rules by accident?

Deletion needs `dryRun: false` and `confirm: true`, and removed items are archived for restore.

Is the graph exposed on the network?

No. It binds to 127.0.0.1 and rejects non-loopback hosts.

Is SyncytiumMD free?

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

How do I add SyncytiumMD to Claude Code?

Run this command in your terminal: claude mcp add syncytiummd -- npx -y syncytium-md

What are the best alternatives to SyncytiumMD?

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

Alternatives to SyncytiumMD

  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