claude-mcp-bridge
MCP server that lets any agent or MCP host (Cursor, Codex, Antigravity, another Claude Code, your own app) delegate to Claude Code running headless (claude -p). Claude is the strongest model here, so the tool surface is deliberately narrow — reasoning, review and autonomous work. Cheap mechanical work (mapping a repo, scanning files) is intentionally left out: it would burn expensive tokens, and Claude already explores natively (Explore subagent + LSP/Grep/Glob) inside delegate. For cheap exploration use a lighter bridge (e.g. cursor-mcp-bridge).
Mirrors the ergonomics of agy-bridge / codex: every tool takes an optional model and effort, returns a session_id, and supports follow_up to continue without resending context.
Tools
| Tool | Purpose | |------|---------| | delegate | Full autonomous task — Claude runs its own loop with tool access in cwd. | | adversarial_review | Strict review of a diff/plan/files: bugs, security, edge cases. Read-only. | | follow_up | Continue a prior session by session_id. |
Every tool accepts: cwd (project root), model (e.g. opus, sonnet, haiku), effort (low \| medium \| high).
Requirements
- Node ≥ 18
claudeCLI installed and authenticated (claudeonce to log in).
Install
git clone https://github.com/JaimeJunr/claude-mcp-bridge.git
cd claude-mcp-bridge
npm install
npm run build
Register in an MCP host
Claude Code: ``bash claude mcp add claude-bridge -s user -- node /abs/path/to/claude-mcp-bridge/dist/index.js ``
Cursor / Codex / any host — add to its mcp.json: ``json { "mcpServers": { "claude-bridge": { "command": "node", "args": ["/abs/path/to/claude-mcp-bridge/dist/index.js"] } } } ``
Configuration (env)
| Var | Default | Meaning | |-----|---------|---------| | CLAUDE_BIN | claude | Path to the claude binary. | | CLAUDE_BRIDGE_PERMISSION | acceptEdits | Permission mode: plan \| default \| acceptEdits \| bypassPermissions. | | CLAUDE_BRIDGE_TIMEOUT_MS | 600000 | Per-call timeout. |
Security:
acceptEditslets delegated Claude edit files autonomously;bypassPermissionsadditionally allows unrestricted shell. Only raise this for trusted workspaces. The review / analyze / search tools instruct Claude to stay read-only regardless.
Develop
npm test # vitest — unit tests for arg building / json parsing
npm run dev # run from source via tsx
License
MIT






