@opti-josh-holt/optimizely-fx-mcp
v1.0.7
Published
An MCP server for Optimizely's Feature Experimentation
Readme
Optimizely Feature Experimentation MCP Server
A Model Context Protocol (MCP) server that provides seamless integration with Optimizely's Feature Experimentation platform, enabling AI assistants to help manage feature flags, audiences, events, and generate SDK implementation code.
🚀 Features
Project Management
- Environment Management: List and retrieve environment details and datafiles
- Collaborator Management: View project collaborators and permissions
- Change History: Track feature flag and project modifications over time
Feature Flag Operations
- Flag Management: List, filter, and retrieve individual feature flags by key
- Flag Status Monitoring: Check flag status across different environments
- Archive Management: Filter flags by archived status
Audience & Targeting
- Audience Management: List and retrieve audience definitions and targeting rules
- Attribute Management: Manage user attributes used for targeting
- Event Tracking: List and retrieve custom events for experimentation
SDK Code Generation
- React Setup: Generate complete React SDK boilerplate with TypeScript support
- Hook Examples: Get working examples for
useDecision,useTrackEvent, anduseFeaturehooks - Installation Guides: Environment-specific installation instructions (browser, Node.js, React, Next.js)
- Initialization Examples: Complete SDK initialization code with various configuration options
- Troubleshooting: Comprehensive debugging guidance and common issue resolution
🔧 MCP Server Configuration
Add the following configuration JSON to your MCP client's config file:
{
"mcpServers": {
"optimizely-fx": {
"command": "npx",
"args": ["-y", "@opti-josh-holt/optimizely-fx-mcp"],
"env": {
"API_ACCESS_TOKEN": "YOUR_API_KEY_HERE"
}
}
}
}Replace API_ACCESS_TOKEN with a REST API token from your profile in Optimizely. After adding the configuration, restart your application (Cursor or Claude Desktop).
💡 Example Prompts
Here are some example prompts you can use with this MCP server:
Feature Flag Management
- "List all feature flags in my project"
- "Show me the details of the 'new_checkout_flow' feature flag"
- "What environments do I have configured?"
- "Get the datafile for my staging environment"
Audience & Targeting
- "Show me all audiences in my project"
- "What attributes are available for targeting?"
- "List all custom events being tracked"
SDK Implementation
- "Generate a React setup for Optimizely with TypeScript"
- "Show me examples of how to use the useDecision hook"
- "How do I install the Optimizely SDK in a Next.js project?"
- "Generate initialization code for a Node.js environment with polling configuration"
- "Help me troubleshoot why my feature flags aren't loading"
Project Insights
- "Show me recent changes to my feature flags"
- "Who are the collaborators on this project?"
- "What's the change history for my project?"
