Model Context Protocol

AI-Powered Error Management

Connect your AI assistant to noisegate using the Model Context Protocol. Manage error patterns, filters, and custom rules through natural language commands.

Get API KeySetup Guide

What is MCP?

The Model Context Protocol is an open standard that enables AI assistants like Claude, ChatGPT, and others to interact with external tools and data sources.

With noisegate's MCP server, you can ask your AI assistant to:

  • Search for error patterns that match your issue
  • Add filters to suppress noisy errors automatically
  • Create custom patterns for your organization
  • Check for duplicates before submitting
Example conversation

You: Find patterns that match "ResizeObserver"

AI: Found 3 patterns matching "ResizeObserver":

  • • ResizeObserver loop limit exceeded (95% confidence)
  • • ResizeObserver loop completed with undelivered notifications (89%)
  • • ResizeObserver is not defined (72%)

You: Enable the first one for my organization

AI: Done! "ResizeObserver loop limit exceeded" is now enabled. Matching errors will be filtered.

Available Tools

9 tools for managing error patterns and filters through your AI assistant

list_patterns

Search and filter error patterns from the catalog

get_pattern

Get detailed info about a pattern with votes and comments

submit_pattern

Submit a new error pattern to the catalog

check_duplicate_patterns

Find similar patterns before submitting

get_filter_config

View your organization's enabled filters

toggle_filter

Enable or disable pattern filters

create_custom_pattern

Create custom patterns for your org

Setup Guide

Get started in minutes with your favorite AI assistant

1

Get your API Key

Navigate to your organization settings and generate an API key. Keep this secure - it authenticates all MCP requests.

2

Configure your AI client

Add the noisegate MCP server to your AI client's configuration:

Claude Desktop (claude_desktop_config.json)

{
  "mcpServers": {
    "noisegate": {
      "url": "https://noisegate.ai/api/mcp",
      "headers": {
        "Authorization": "Bearer YOUR_API_KEY"
      }
    }
  }
}

OpenAI ChatGPT (via MCP bridge)

{
  "mcpServers": {
    "noisegate": {
      "transport": "streamable-http",
      "url": "https://noisegate.ai/api/mcp",
      "headers": {
        "Authorization": "Bearer YOUR_API_KEY"
      }
    }
  }
}

Cursor (.cursor/mcp.json)

{
  "mcpServers": {
    "noisegate": {
      "url": "https://noisegate.ai/api/mcp",
      "headers": {
        "Authorization": "Bearer YOUR_API_KEY"
      }
    }
  }
}

Windsurf (mcp_config.json)

{
  "mcpServers": {
    "noisegate": {
      "serverUrl": "https://noisegate.ai/api/mcp",
      "headers": {
        "Authorization": "Bearer YOUR_API_KEY"
      }
    }
  }
}

Generic HTTP Configuration

Endpoint: https://noisegate.ai/api/mcp
Method: POST (for requests), GET (for SSE streams)
Headers:
  Authorization: Bearer YOUR_API_KEY
  Content-Type: application/json
3

Start using noisegate

Once connected, you can interact with noisegate through natural language. Try these commands:

  • "Search for patterns related to browser extensions"
  • "Show me my organization's current filter config"
  • "Create a custom pattern for our internal error message"
  • "Find duplicates before I submit this pattern"

Ready to get started?

Generate your API key and connect your AI assistant to noisegate in minutes.

Generate API KeyMCP Documentation