unleash-ff
Unleash MCP Server
claude mcp add --transport stdio ylin6-unleash-ff-mcp-server npx -y @ylin6/unleash-ff-mcp-server \ --env UNLEASH_API_URL="your-unleash-api-url" \ --env UNLEASH_AUTH_TOKEN="your-unleash-auth-token"
How to use
This MCP server provides a bridge between AI agents and your Unleash feature flag management system. It exposes a set of tools that let you list projects, enumerate feature flags within a project, and create, update, or fetch details for individual feature flags. Use these tools to programmatically manage feature flags from within your agent workflows, enabling dynamic behavior controlled by feature flags in Unleash.
To run locally, ensure you have a valid Unleash instance and supply the required environment variables (UNLEASH_API_URL and UNLEASH_AUTH_TOKEN). The server is intended to be invoked via npx (or npm install -g) and can be used with clients like Cursor or other MCP-enabled assistants. The available operations include: Get Projects (list all projects), Get Features (list all flags within a project), Create Feature Flag (create a new flag with a name, description, and type), Update Feature Flag (update description or type), and Get Feature Flag (retrieve details for a specific flag). These tools enable automated experimentation and feature rollout planning from AI workflows.
How to install
Prerequisites:
- Node.js and npm installed on your system
- Access to a Unleash instance with API URL and a valid auth token
Installation options:
Option 1: Install globally via npm
npm install -g @ylin6/unleash-ff-mcp-server
Option 2: Run directly with npx (no local install required)
npx -y @ylin6/unleash-ff-mcp-server
Option 3: Use within a project (local dev/testing)
npm install @ylin6/unleash-ff-mcp-server --save-dev
Configuration (environment variables):
- UNLEASH_API_URL: Your Unleash API URL
- UNLEASH_AUTH_TOKEN: Authentication token for Unleash
Run example (with npx):
env UNLEASH_API_URL=https://your-unleash-url/api UNLEASH_AUTH_TOKEN=token npx -y @ylin6/unleash-ff-mcp-server
Additional notes
Tips:
- Ensure your Unleash API URL is reachable from where you run the MCP server.
- The MCP server relies on environment variables for authentication; never hard-code tokens in shared code.
- If you encounter authentication or CORS-related errors from Unleash, verify token scopes and API accessibility.
- Use the provided MCP inspector (npm run inspect) to analyze request/response patterns and refine prompts or tool usage.
- When using with Cursor or other agents, set the environment variables in the agent configuration alongside the npx command.
Common issues:
- Missing UNLEASH_API_URL or UNLEASH_AUTH_TOKEN leads to authentication failures.
- Network/firewall restrictions blocking the Unleash API endpoint.
- Incorrect projectId or featureId values in tool parameters.
Configuration options:
- You can extend the environment block with additional vars if the MCP server exposes them in the runtime.
- Consider wrapping the server in a container for consistent environments acrossDevelopment and production.
Related MCP Servers
iterm
A Model Context Protocol server that executes commands in the current iTerm session - useful for REPL and CLI assistance
mcp
Octopus Deploy Official MCP Server
furi
CLI & API for MCP management
editor
MCP Server for Phaser Editor
DoorDash
MCP server from JordanDalton/DoorDash-MCP-Server
mcp
MCP сервер для автоматического создания и развертывания приложений в Timeweb Cloud