OpenFeature FAQs

Question 1

What is the OpenFeature MCP Local Server?

Accepted Answer

It's a local Model Context Protocol (MCP) server designed to provide OpenFeature SDK installation guidance and facilitate remote feature flag evaluation using the OpenFeature Remote Evaluation Protocol (OFREP) over stdio, primarily for local AI client integration.

Question 2

How does OpenFeature MCP help with SDK installation?

Accepted Answer

It delivers dynamic, technology-specific Markdown instructions for installing various OpenFeature SDKs (e.g., Node.js, Python, Java). It can also augment these guides with provider-specific installation notes from the OpenFeature ecosystem.

Question 3

Can I integrate this tool with my AI development environment?

Accepted Answer

Yes, it's designed for local integration with MCP-compatible AI clients like Cursor, VS Code (Continue), Claude Code, Windsurf, and Claude Desktop, using `npx @openfeature/mcp` as a command within their `mcpServers` configuration.

Question 4

How can I configure authentication for OFREP flag evaluations?

Accepted Answer

Authentication for OFREP can be configured with a bearer token or API key. These can be provided directly in tool call arguments, via environment variables (`OPENFEATURE_OFREP_BEARER_TOKEN`, `OPENFEATURE_OFREP_API_KEY`), or within a `~/.openfeature-mcp.json` config file.

Question 5

What is OFREP, and how does this tool utilize it?

Accepted Answer

OFREP (OpenFeature Remote Evaluation Protocol) is used by this tool to perform single or bulk feature flag evaluations against a compatible flag service. It supports flexible configuration for `base_url` and authentication (bearer token, API key) via args, environment variables, or a config file.

OpenFeature

About

Key Features

Use Cases