Skip to content
MCP server

Rush

Hand coding tasks to cloud agents from your client and collect the PR, report or files

Install
Pricing
Freemium
Platforms
API, Web, CLI
Runs
Remotely
Version
1.2.0
Listing updated
3 days ago

About Rush

Rush is a freemium MCP server for developer tools. It runs as a hosted remote server, so there is nothing to install locally.

Work runs elsewhere: each run gets its own cloud sandbox with a computer and a browser, and the client only starts it and reads the result. create_session returns immediately with a session id, and wait_for_sessions waits up to 50 seconds per call, so an agent can launch a batch of runs in parallel and collect them as they finish. A webhook on the key is the alternative for a server that should hear back without holding a call open.

Tools cover sessions (create, get, list, message, cancel, events, changes, artifacts), saved agents, projects, models, connections and repositories. If a run is refused because an account is not connected, connect_account returns a sign-in link for the person to open, so no secret passes through the model. Claude Code, Codex or OpenCode is picked per run.

Claude.ai and ChatGPT connect through OAuth, and an admin of the organisation must approve. Claude Code, Codex, Cursor and Notion Custom Agents can use an API key. A plugin adds six skills such as fix-a-bug and review-a-pr. Pricing is per member for cloud hours; Free gives 2 agent-hours a month.

Install

Run in a terminal

claude mcp add --transport http rush https://api.phoenix.so/mcp

Add to Settings → Connectors → Add custom connector

https://api.phoenix.so/mcp
Install in CursorOpens Cursor, which asks you to confirm. Or add it by hand:

Add to ~/.cursor/mcp.json

{
  "mcpServers": {
    "rush": {
      "url": "https://api.phoenix.so/mcp"
    }
  }
}
Install in VS CodeOpens VS Code, which asks you to confirm. Or add it by hand:

Add to .vscode/mcp.json

{
  "servers": {
    "rush": {
      "type": "http",
      "url": "https://api.phoenix.so/mcp"
    }
  }
}

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

{
  "mcpServers": {
    "rush": {
      "serverUrl": "https://api.phoenix.so/mcp"
    }
  }
}

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

{
  "mcpServers": {
    "rush": {
      "type": "http",
      "url": "https://api.phoenix.so/mcp"
    }
  }
}

Tools this server provides (19)

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

create_session
Starts a run and returns the session id.
get_session
Reads one session.
wait_for_sessions
Waits up to 50 seconds for sessions to settle.
list_sessions
Lists sessions.
send_session_message
Sends a follow-up to a run.
cancel_session
Cancels a run.
list_session_events
Lists events of a session.
get_session_changes
Reads the changes of a session.
get_session_artifact
Reads an artifact of a session.
list_agents
Lists saved agents.
get_agent
Reads a saved agent.
create_agent
Creates a saved agent.
update_agent
Updates a saved agent.
list_projects
Lists projects.
get_project
Reads a project.
list_models
Lists models available to a run.
list_connections
Lists connected accounts.
connect_account
Starts account sign-in and returns an authorize link.
list_repositories
Lists repositories a run can use.

What it does

Session control
Starts, monitors, messages and cancels cloud runs as tool calls.
Parallel batches
wait_for_sessions lets one agent start several runs and collect them as they settle.
Saved agents
Lists, creates and updates saved agents and reads projects.
Isolated sandboxes
Each run has its own cloud computer and browser.
Safe account linking
connect_account returns a link, so no token passes through the model.
Plugin skills
Six skills include fix-a-bug, review-a-pr and run-in-parallel.

Good for

  • Hand a bug fix to a cloud agent and get a pull request
  • Review a PR in a separate sandbox
  • Run several tasks in parallel from one chat
  • Save a repeatable agent for recurring work
  • Pull a report or files produced by a run

Pricing

Freemium, from $40 a month

Free 2 agent-hours/mo; Pro $40/member/mo; Max $300/member/mo

Prices change. Check getrush.ai before you buy.

Rush FAQ

Does the run happen on my machine?

No. Each run has its own cloud sandbox; your client sends the request and reads the result.

How do I sign in?

OAuth from claude.ai or ChatGPT, or an API key from the console.

What does a run cost?

Plans pay for cloud agent-hours; models are billed separately after the first $1, and runs from OAuth sign-in use your own account.

How long does a run take?

Minutes, so the tools return an id and a wait call collects results.

Is Rush free?

Rush has a free plan, with paid plans for more usage or features. Paid plans start at $40 a month.

How do I add Rush to Claude Code?

Run this command in your terminal: claude mcp add --transport http rush https://api.phoenix.so/mcp

What are the best alternatives to Rush?

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

Alternatives to Rush

  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