home / mcp / honeybadger mcp server

Honeybadger MCP Server

Bridges Honeybadger error data to MCP clients, enabling list and detail queries on faults.

Installation

Add the following to your MCP client configuration file.

Configuration

{
  "mcpServers": {
    "bobtista-honeybadger-mcp": {
      "command": "uv",
      "args": [
        "run",
        "src/honeybadger_mcp_server/server.py",
        "--transport",
        "stdio",
        "--api-key",
        "YOUR-API-KEY",
        "--project-id",
        "YOUR-PROJECT-ID"
      ],
      "env": {
        "TRANSPORT": "stdio",
        "HONEYBADGER_API_KEY": "YOUR-API-KEY",
        "HONEYBADGER_PROJECT_ID": "YOUR-PROJECT-ID"
      }
    }
  }
}

This MCP server bridges the Honeybadger error monitoring service with AI agents, enabling you to fetch and analyze fault data from your Honeybadger projects. It exposes reliable endpoints that let clients list faults and drill into fault details, supporting efficient error investigation and automation.

How to use

You connect an MCP client to the Honeybadger MCP Server using a local or containerized stdio workflow. Start the server with the stdio transport so your client can spin up and communicate directly. You can run the server with your API key and project ID, then use the client to list faults and fetch fault details as needed. The two main capabilities are: listing faults with optional text, time, sorting, and paging; and retrieving detailed fault notices with time-based filters and paging.

How to install

Prerequisites you need before installing: Python 3.10 or newer, a Honeybadger API key, and your Honeybadger Project ID. Docker is recommended if you want to run the MCP server inside a container.

Configuration and running options

Choose your preferred run method and provide your API credentials and project information. The server supports both environment-variable based startup and explicit command-line startup.

{
  "mcpServers": {
    "honeybadger": {
      "transport": "stdio",
      "command": "uv",
      "args": [
        "run",
        "src/honeybadger_mcp_server/server.py",
        "--transport",
        "stdio",
        "--api-key",
        "YOUR-API-KEY",
        "--project-id",
        "YOUR-PROJECT-ID"
      ]
    }
  }
}

Using the installed package or containers

You can start the MCP server directly if you have the package installed or you run it inside a container.

# Using the installed package (stdio)
honeybadger-mcp-server --transport stdio --api-key YOUR-API-KEY --project-id YOUR-PROJECT-ID

# Using Docker (stdio, example)
docker run --rm -i -e HONEYBADGER_API_KEY=YOUR-API-KEY -e HONEYBADGER_PROJECT_ID=YOUR-PROJECT-ID -p 8050:8050 honeybadger/mcp --transport stdio --api-key YOUR-API-KEY --project-id YOUR-PROJECT-ID

Connecting with an MCP client

Configure your client with the provided server information and use the available tools to list faults and view fault details. The client will use the chosen transport to communicate with the server and perform queries against your Honeybadger data.

Available tools

list_faults

Lists and filters faults from your Honeybadger project with optional text search, time filters, and sorting options. Supports pagination.

get_fault_details

Retrieves detailed information about specific faults, with options to filter notices by creation time and paginate results. Results are ordered by creation time descending.