Fastly MCP brings the power of Fastly's API directly to your AI assistants through the Model Context Protocol (MCP).
Fastly's API-first design philosophy means:
- Everything is an API - Every feature available in the Fastly UI is accessible via API
- Programmatic Control - Full control over services, configurations, and edge logic
- Automation Ready - Support for CI/CD workflows and infrastructure as code
- Real-time Changes - API changes propagate globally in seconds, not minutes or hours
Fastly's comprehensive API allows you to:
- Manage CDN Services - Create, configure, and deploy content delivery services
- Control Caching - Set up cache strategies and perform instant purges
- Configure Security - Manage WAF, DDoS protection, and TLS certificates
- Monitor Performance - Access real-time metrics and historical stats
- Implement Edge Logic - Deploy custom VCL or Compute@Edge applications
- Automate Workflows - Integrate with CI/CD pipelines and infrastructure tools
The AI assistant never sees your Fastly API key. It talks to a local helper (FastlyMCP) which uses the key securely.
With Fastly MCP configured, you can ask your AI assistant questions like:
| What You Want To Do | Example AI Request |
|---|---|
| List your services | "Show me all my Fastly services" |
| Get domain details | "What domains are configured for my e-commerce service?" |
| Purge cache | "Purge the cache for my product service" |
| Check traffic | "What's the traffic pattern for my main site over the last week?" |
| View configuration | "Show me the backend servers for my API service" |
| Check performance | "What's my current cache hit ratio?" |
- A Fastly account and API key (Get started with Fastly)
- An AI assistant that supports MCP (e.g., Claude, GPT with plugins)
- The Fastly CLI installed (Installation Guide)
Configure your AI assistant with:
{
"mcpServers": {
"fastly": {
"command": "node",
"args": ["path/to/fastly-mcp.mjs"],
"env": {
"FASTLY_API_KEY": "your_fastly_api_key"
}
}
}
}| Task Goal | Example AI Request |
|---|---|
| Optimize service based on traffic | "Analyze the configuration for [service_id/name] and suggest optimizations based on its traffic profile, prioritizing low latency." |
| Configure for live video | "Configure [service_id/name] for optimal live video streaming following the best practices outlined in [link_to_guide_or_doc]." |
| Find config conflicts | "Identify potential configuration conflicts in [service_id/name] compared to standard e-commerce delivery patterns." |
| Optimize video chunk caching | "Optimize caching for [service_id/name] to handle 10-second video chunks efficiently, minimizing origin load." |
| Enhance WAF security | "Review the WAF rules for [service_id/name] and suggest stricter settings to mitigate potential SQL injection attacks." |
| Set up origin mTLS | "Set up Mutual TLS (mTLS) authentication between Fastly and the origin servers for [service_id/name]." |
| Implement A/B testing (Edge) | "Deploy a Compute@Edge function to [service_id/name] that performs A/B testing by routing 10% of users to backend [backend_name]." |
| Add dynamic image rewriting (VCL) | "Write and deploy VCL for [service_id/name] to dynamically rewrite image URLs based on the requesting device's user agent." |
| Troubleshoot 5xx errors | "Analyze logs for [service_id/name] from the past 24 hours to identify the root cause of the recent spike in 5xx errors." |
This project is licensed under the MIT License - see the LICENSE file for details.