Fetches and analyzes test run data from the Sippy API to facilitate AI-driven regression and failure analysis.
The Fetch Test Runs skill integrates Claude with the Sippy API to retrieve detailed test execution data, including JUnit outputs and failure logs. It is specifically designed to support AI-based similarity analysis, helping developers determine if multiple test failures share a common root cause or are isolated incidents. By providing features like job name filtering and mass failure detection, it streamlines the process of debugging complex CI pipelines and identifying infrastructure-related regressions in OpenShift and Kubernetes environments.
Key Features
01Retrieves raw JUnit test failure outputs for deep AI-based similarity analysis
02Supports both JSON and human-readable summary output formats
03Filters runs by job name substrings and specific Prow job run IDs
0460 GitHub stars
05Configurable lookback periods and the ability to include successful runs
06Identifies mass failure jobs to distinguish between code bugs and infrastructure issues
Use Cases
01Filtering CI data to specific cloud providers or technical preview environments
02Automated root cause analysis of flaky or failing CI tests
03Comparing error messages across different jobs to find recurring regression patterns
