Skip to main content

MCP Support

Polar Signals provides a Model Context Protocol (MCP) server that enables AI assistants to connect directly with performance profiling data sources. This transforms performance analysis by allowing natural language queries about application performance data.

What is MCP?​

The Model Context Protocol enables AI assistants to securely connect to external data sources and tools. With Polar Signals' MCP server, you can:

  • Analyze performance data using natural language: Ask questions like "What are the main CPU bottlenecks?" or "Show memory allocation patterns"
  • AI-driven code optimization: Get specific suggestions for improving performance based on profiling data
  • Explore profiling data: Discover available profile types, labels, and values in your projects

Setup​

1. Generate an API Token​

First, generate an API token in your Polar Signals project settings.

2. Add the MCP Server​

Using Claude Code CLI​

Setting up Remote MCP with Polar Signals Cloud is straightforward:

claude mcp add --transport http polarsignals https://api.polarsignals.com/api/mcp/ --header "Authorization: Bearer YOUR_TOKEN"

No extra installation is required. Your LLM can directly talk to our Remote MCP.

Using Cursor​

Add this configuration to your Cursor settings:

{
  "mcpServers": {
    "polarsignals": {
      "url": "https://api.polarsignals.com/api/mcp/",
      "headers": {
        "Authorization": "Bearer YOUR_TOKEN"
      }
    }
  }
}

3. Remove the MCP Server​

claude mcp remove polarsignals

Usage Examples​

Once configured, you can ask natural language questions about your profiling data:

  • "What are the main CPU bottlenecks in my application?"
  • "Show me memory allocation patterns for the last hour"
  • "Which functions consume the most CPU time?"
  • "Analyze the performance profile and suggest optimizations"

The MCP server will translate your questions into precise profile queries and provide actionable insights for performance optimization.

Available Tools​

The Polar Signals MCP server provides the following tools:

  • profile_types: Discover available profile types in your project
  • labels: Explore metadata labels in profiles
  • values: Get specific label values for filtering
  • get_profile: Query profiles using PromQL syntax