An unofficial TypeScript MCP server for BugBug, a test automation platform, that lets an AI assistant manage the full test lifecycle: browsing and editing tests and suites, triggering a run, waiting for it to complete, and getting the full results or an explanation of a failure, without switching to the BugBug web dashboard. Test execution is smart rather than requiring an exact identifier - a test can be run by name or UUID with intelligent matching - and completion can be waited on synchronously so the assistant returns actual results instead of just a run-started acknowledgment.

Use it when an AI assistant needs to trigger a BugBug test or suite and act on the outcome - running a specific test by name, checking what ran in the last 24 hours, or getting an explanation of why a test failed. It works across AI assistants beyond Claude, including Windsurf and GitHub Copilot, since it speaks the standard MCP protocol. It requires an active BugBug account with an API token; without one, none of the tools can authenticate. Because this is an unofficial, community-built implementation rather than one maintained by BugBug itself, it is worth treating with the same caution as any third-party wrapper around a paid platform's API - review what it does with the API token before granting it access to a production test suite.

Execution and monitoring: run_test_by_name_or_id, wait_for_test_run, wait_for_suite_run, which block until completion and return full results, show_run_from_last_24, explain_error. Tests and suites: get_tests, get_test, update_test (full update), partial_update_test, get_suites, get_suite. Runs and profiles: get_test_runs, get_test_run, get_profiles, get_profile.

npx -y @simplypixi/bugbug-mcp-server@latest

Configure the AI assistant's MCP settings:

{
  "mcpServers": {
    "bugbug": {
      "command": "npx",
      "args": ["-y", "@simplypixi/bugbug-mcp-server@latest"],
      "env": {
        "API_KEY": "your_bugbug_api_token_here"
      }
    }
  }
}

Requires Node.js 18 or higher and a BugBug account with an API token.

QA and development teams using BugBug who want an AI assistant to trigger tests, wait on and interpret results, and manage suites and profiles directly from a conversation.