Retrieve the authenticated user's recent Evergreen patches/commits with their CI/CD status. Use this to see your recent code changes, check patch status (success/failed/running), and identify patches that need attention. Returns patch IDs needed for other tools. If project_id is not specified, will automatically detect it from your workspace directory and recent patch activity.This tool may return a list of available project_ids if it cannot determine the project_id automatically.You should ask the user which project they want to use, then call this tool again with the project_id parameter set to their choice.

Analyze failed CI/CD jobs for a specific patch to understand why builds are failing. Shows detailed failure information including failed tasks, build variants, timeout issues, log links, and test failure counts. Essential for debugging patch failures. If project_id is not specified, will automatically detect it from your workspace directory and recent patch activity.This tool may return a list of available project_ids if it cannot determine the project_id automatically.You should ask the user which project they want to use, then call this tool again with the project_id parameter set to their choice.

Get a truncated view of task logs via GraphQL. Returns log metadata and filtered error/failure messages, but only captures a limited portion of the full log (mostly test log ingestion messages). For complete raw task logs including timeout output, process dumps, and full execution logs, use get_task_log_detailed instead. Use task_id from get_patch_failed_jobs results.

Get test result metadata via GraphQL. Returns test names, pass/fail statuses, durations, and Parsley log viewer URLs — but not the actual error messages from test output. For the raw test log content with error pattern analysis, use get_test_results_detailed instead. Use task_id from get_patch_failed_jobs results.

Get a list of unique project identifiers inferred from the user's recent patches. This helps discover which Evergreen projects the user has been working on, sorted by activity (patch count and recency). Useful for understanding project context and filtering other queries.

Get the complete raw task logs via REST API. Returns the full untruncated task execution log including timeout handler output, process dumps, and stdout/stderr — content that the GraphQL get_task_log_summary tool cannot access. Automatically scans for error patterns and returns a structured summary with top error terms and example lines when errors are found. Best for debugging non-test failures (setup errors, timeouts, compilation failures). Use task_id from get_patch_failed_jobs results.

Get raw test log content via REST API. Fetches actual test output (stored in S3, not accessible via GraphQL). Automatically scans for error patterns and returns a structured summary with top error terms and example lines when errors are found. Use this to understand WHY a test failed, not just that it failed. Requires task_id and test_name from get_patch_failed_jobs results.

Download artifacts from a specific Evergreen task. Use this to retrieve build outputs, test results, logs, or other files generated by a task. Artifacts are downloaded to a local directory structure organized by version.