OpenFeature MCP Server
The OpenFeature MCP Server is a Cloudflare Worker-based service that provides OpenFeature SDK installation guidance via the MCP protocol. It offers an install_openfeature_sdk tool to fetch Markdown-based installation guides (currently supporting React SDK), supports MCP communication over both JSON-RPC (/mcp) and Server-Sent Events (/sse) transports, and includes health check (/health) and service information (/info) endpoints. The server requires no authentication and leverages Cloudflare Workers for global edge distribution and serverless scalability.
Provides OpenFeature SDK installation guidance and setup instructions for React applications
Click on "Install Server".
Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state.
In the chat, type
@followed by the MCP server name and your instructions, e.g., "@OpenFeature MCP Servershow me how to install the OpenFeature SDK for React with LaunchDarkly"
That's it! The server will respond to your query, and you can continue using it as needed.
Here is a step-by-step guide with screenshots.
Archived: MIGRATED TO OPENFEATURE ORG
OpenFeature MCP Local Server (stdio)
Warning
This project is in active development.
Related MCP server: OpenFeature MCP Server
Features
A local Model Context Protocol (MCP) server that provides OpenFeature SDK installation guidance and Open Feature Remote Evaluation Protocol (OFREP) over stdio.
OpenFeature SDK Installation Guides: Fetch installation prompts for various OpenFeature SDKs
MCP stdio Transport: Intended for local usage by MCP-compatible clients
Configure your AI client (local)
Cursor
Add to ~/.cursor/mcp_settings.json:
{
"mcpServers": {
"OpenFeature": {
"command": "npx",
"args": ["-y", "@openfeature/mcp"]
}
}
}VS Code (Continue)
Add to .continue/config.json:
{
"mcpServers": {
"OpenFeature": {
"command": "npx",
"args": ["-y", "@openfeature/mcp"]
}
}
}Claude Code (CLI)
Add the server via CLI:
claude mcp add --transport stdio openfeature npx -y @openfeature/mcpThen manage the connection in the CLI with /mcp.
Windsurf
In the "Manage MCP servers" raw config, add:
{
"mcpServers": {
"OpenFeature": {
"command": "npx",
"args": ["-y", "@openfeature/mcp"]
}
}
}Claude Desktop
Edit your Claude Desktop config and add:
{
"mcpServers": {
"openfeature": {
"command": "npx",
"args": ["-y", "@openfeature/mcp"]
}
}
}Restart Claude Desktop after saving.
NPM Global install (optional)
If you prefer a global install instead of NPX:
npm install -g @openfeature/mcpNow in your MCP config use openfeature-mcp as the command:
{
"mcpServers": {
"openfeature": {
"command": "openfeature-mcp"
}
}
}All logs are written to stderr. The MCP protocol messages use stdout.
Available Tools
install_openfeature_sdk
Fetches Markdown instructions for installing the OpenFeature SDK for a given technology. Optionally augments the prompt with installation guidance for one or more feature flag providers.
Parameters:
technology(string enum): One of the supported technologies listed belowproviders(string array, optional): Zero or more provider identifiers. If present, adds provider-specific installation notes to the prompt (or removes placeholder sections when empty).
Supported Technologies:
The technologies list is build from the avaliable prompts/*.md, updated automatically using scripts/build-prompts.js
android
dotnet
go
ios
java
javascript
nestjs
nodejs
php
python
react
ruby
Supported Providers:
The provider list is sourced automatically from the OpenFeature open-feature/openfeature.dev
repo; newly added providers there become available here without manual edits.
See scripts/build-providers.js for details.
ofrep_flag_eval
Evaluate feature flags via OpenFeature Remote Evaluation Protocol (OFREP).
If flag_key is omitted, performs bulk evaluation.
References:
open-feature/protocol repo,
OFREP OpenAPI spec
Parameters (all optional unless noted):
base_url(string, optional): Base URL of your OFREP-compatible flag service. If omitted, the server uses env/config (see below).flag_key(string, optional): If provided, calls single flag evaluation:/ofrep/v1/evaluate/flags/{key}. If omitted, calls bulk:/ofrep/v1/evaluate/flags.context(object, optional): Evaluation context, e.g.{ "targetingKey": "user-123", ... }.etag(string, optional): For bulk requests, sent asIf-None-Matchto enable 304 caching semantics.auth(object, optional): Inline auth for this call only.bearer_token(string, optional): SetsAuthorization: Bearer <token>.api_key(string, optional): SetsX-API-Key: <key>.
Auth and base URL resolution (priority):
Tool call args:
base_url,auth.bearer_token,auth.api_keyEnvironment variables:
OPENFEATURE_OFREP_BASE_URL(orOFREP_BASE_URL),OPENFEATURE_OFREP_BEARER_TOKEN(orOFREP_BEARER_TOKEN),OPENFEATURE_OFREP_API_KEY(orOFREP_API_KEY)Config file:
~/.openfeature-mcp.json(override withOPENFEATURE_MCP_CONFIG_PATH)
Example ~/.openfeature-mcp.json:
{
"OFREP": {
"baseUrl": "https://flags.example.com",
"bearerToken": "<token>",
"apiKey": "<key>"
}
}Notes:
Bulk requests may return
ETag. Pass it back viaetagto leverage 304 Not Modified.Either bearer token or API key can be supplied; both are supported by the spec.
Development
Prerequisites
Node.js 18+
Setup
Install dependencies:
npm installAdd or edit install guides in the
prompts/folder (Markdown). These are bundled at build time.Build prompts bundle:
npm run build-promptsBuild TypeScript:
npm run buildRun locally (binary entrypoint):
node dist/cli.js
Available Tools
1 toolinstall_openfeature_sdkB
Fetch and return OpenFeature install prompt Markdown by guide name. Available guides: android, dotnet, go, ios, java, javascript, nestjs, nodejs, php, python, react, ruby. Input: { guide: string }
| Name | Required | Description | Default |
|---|---|---|---|
| guide | Yes |
TDQS
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
No annotations are provided, so the description must fully disclose behavioral traits. It mentions fetching and returning Markdown, but lacks details on permissions, rate limits, error responses, or whether the operation is read-only or has side effects. For a tool with no annotation coverage, this leaves significant gaps in understanding its behavior and constraints.
Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.
Is the description appropriately sized, front-loaded, and free of redundancy?
The description is highly concise and front-loaded, with two sentences that efficiently convey the tool's purpose and input details. Every word serves a purpose, avoiding redundancy or unnecessary elaboration, making it easy to parse quickly.
Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.
Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?
Given the tool's low complexity (1 parameter with enum, no output schema, no annotations), the description is minimally adequate. It covers the basic purpose and parameter options but lacks details on output format, error handling, or behavioral traits. Without annotations or an output schema, more context on what the Markdown contains or how to handle failures would improve completeness.
Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.
Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?
The description adds meaningful context beyond the input schema by listing all available guide options (e.g., android, dotnet, go) and specifying that the input is a guide name. Since the schema description coverage is 0% (no descriptions in schema properties), the description compensates well by clarifying parameter semantics, though it could detail the format or purpose of the returned Markdown more explicitly.
Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.
Does the description clearly state what the tool does and how it differs from similar tools?
The description clearly states the tool's purpose: 'Fetch and return OpenFeature install prompt Markdown by guide name.' It specifies the verb ('fetch and return'), resource ('OpenFeature install prompt Markdown'), and scope ('by guide name'), making the action concrete. However, since there are no sibling tools mentioned, it cannot differentiate from alternatives, preventing a perfect score.
Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.
Does the description explain when to use this tool, when not to, or what alternatives exist?
The description implies usage by listing available guides, suggesting this tool is used to retrieve installation instructions for specific platforms. However, it does not provide explicit guidance on when to use this tool versus alternatives, prerequisites, or error handling. With no sibling tools, the context is limited, but more detailed instructions could enhance clarity.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
TDQS
With only one tool, there is no possibility of ambiguity or overlap between tools. The tool has a single, clear purpose: fetching OpenFeature SDK installation prompts by guide name.
Since there is only one tool, naming consistency is inherently perfect. The tool name 'install_openfeature_sdk' follows a clear verb_noun pattern and is descriptive of its function.
A single tool is too few for a server named 'OpenFeature MCP Server', which suggests a broader scope related to feature flag management. This minimal set severely limits functionality and agent workflows, making it feel incomplete and under-scoped.
The tool set is severely incomplete for the apparent domain of OpenFeature (feature flag management). It lacks core operations like creating, updating, or evaluating feature flags, managing providers, or accessing flag configurations, leaving significant gaps that will cause agent failures.
Maintenance
Resources
Unclaimed servers have limited discoverability.
Looking for Admin?
If you are the server author, to access and configure the admin panel.
Related MCP Connectors
Discover open-source AI agents and get source-owned installation instructions through MCP.
Connect MCP clients to 2,000+ AI models without managing provider API keys.
Package intelligence MCP for AI agents — 22 tools, 19 ecosystems, AGPL SDK, free.
A live, curated feed of new AI agent capabilities across MCP, SDKs, models, and APIs.
Related MCP Servers
- AlicenseNot gradedqualityCmaintenanceEnables access to agent instruction files and prompts for AI development workflows. Provides tools to retrieve and list development rules, security checks, and common prompts from an agents library through MCP protocol.3Apache 2.0
- AlicenseNot gradedqualityCmaintenanceProvides OpenFeature SDK installation guidance for various programming languages and enables feature flag evaluation through the OpenFeature Remote Evaluation Protocol (OFREP). Supports multiple AI clients and can connect to any OFREP-compatible feature flag service.1513Apache 2.0
- AlicenseNot gradedqualityDmaintenanceProvides MCP tools to list and search OpenAI Agents SDK documentation, enabling LLMs to retrieve documentation topics and content via natural language queries.MIT
- AlicenseNot gradedqualityDmaintenanceMCP server for Langfuse Prompt Management that enables listing and retrieving Langfuse prompts via MCP prompts and tools.MIT
Appeared in Searches
Latest Blog Posts
- Who's Calling? MCP Hosts Are an Identity Blind Spot (And the Spec Knows It)By Om-Shree-0709 on .mcpAgent IdentityOAuth 2.1
- Your AI Chatbot Just Exposed Your CEO's Salary to an InternBy Om-Shree-0709 on .Agent IdentityMCP SecurityOAuth Delegation
- Why MCP Servers Need Execution Sandboxing (And Why Your Current Stack Isn't Enough)By Om-Shree-0709 on .Agentic AiPrompt InjectionWebAssembly
MCP directory API
We provide all the information about MCP servers via our MCP API.
curl -X GET 'https://glama.ai/api/mcp/v1/servers/DevCycleHQ-Sandbox/openfeature-mcp'
If you have feedback or need assistance with the MCP directory API, please join our Discord server