Skip to content
Unleash

UnleashMCP Server

by Unleash
Released
4 views
v0.4.1
Repository Docs npm
Developer Tools

Summary

Unleash's official MCP server for feature flags — and unusually, it first judges whether a change needs a flag at all before creating one.

Features

  • evaluate_change scores the risk of a change and recommends whether a flag is warranted
  • detect_flag finds existing flags so agents reuse rather than duplicate
  • create_flag creates a validated, properly typed flag
  • wrap_change explains how to wrap the code change in the flag
  • set_flag_rollout and toggle_flag_environment control rollout strategy and per-environment state
  • cleanup_flag generates instructions for safely removing flagged code paths
  • list_flags and list_projects with pagination; get_flag_state for metadata and strategies

Installation

Set up this MCP server in your favorite AI agent — copy a ready-made configuration below.

Any MCP-compatible agent

Most agents (Claude, Cursor, Windsurf, VS Code, and more) read a standard mcpServers configuration.

  1. Open your agent's MCP configuration file.
  2. Merge the snippet below into it, filling in the environment variables with your own values.
  3. Restart the agent — the "Unleash" tools will be available.
{
  "mcpServers": {
    "unleash-mcp": {
      "command": "npx",
      "args": [
        "-y",
        "@unleash/mcp"
      ],
      "env": {
        "UNLEASH_BASE_URL": "https://<your-instance>/api",
        "UNLEASH_PAT": "<personal-access-token>"
      }
    }
  }
}

Description

Most feature-flag integrations give an agent a create_flag call and leave the judgement to
the model, which is how codebases end up littered with flags nobody needed. Unleash's official MCP server is built
around the opposite instinct: the intended first call is evaluate_change, which scores the risk of a change and
recommends whether a flag is warranted at all. Only then does detect_flag check whether a suitable flag already
exists, so an agent reuses one instead of minting a near-duplicate.

From there the workflow is complete in both directions. create_flag creates a validated, properly typed flag;
wrap_change explains how to wrap the code change around it; set_flag_rollout configures a rollout strategy
without enabling the flag, and toggle_flag_environment enables or disables it per environment — a separation that
matters, because "configured" and "live" are different states. get_flag_state surfaces a flag's metadata and
activation strategies, and list_flags and list_projects page through what exists. Crucially, cleanup_flag
generates instructions for safely removing flagged code paths, closing the loop that most teams leave open.

It runs locally over stdio from the @unleash/mcp npm package and talks to the Unleash Admin API, so it works
against both hosted and self-managed instances. Configuration needs UNLEASH_BASE_URL and a personal access token
in UNLEASH_PAT, with an optional UNLEASH_DEFAULT_PROJECT. The guidance it encodes comes from Unleash's own
documentation on using feature flags at scale.

Related MCP Servers

MCP: Mux MCP Server

by Mux

New

Official MCP server for the Mux video API, built on a code-execution scheme: the agent writes TypeScript against the SDK and runs it in a Deno sandbox.

Developer Tools
8 views

MCP: ripwire

by Red Hat

Featured

Red Hat's zero-dependency C++23 code-context engine — ranked call graphs and blast-radius analysis, indexing a repo in under half a second with no server and no database.

Developer Tools
9 views

MCP: Graft

by Trail

Featured

Builds a searchable markdown graph of your repo so coding agents stop re-exploring it on every task — reported 42% fewer tokens and 46% fewer tool calls.

Developer Tools
8 views

MCP: Expo MCP Server

by Expo

Featured

Expo's official remote MCP server — searches Expo docs, installs compatible SDK packages, triggers and monitors EAS builds, and drives iOS/Android simulators.

Developer Tools
7 views
Browse all MCP servers →