What problem does it solve? CI/CD pipelines fail for opaque reasons, and developers waste cycles pushing commits just to see if a fix works on remote runners. This Skill autonomously reproduces pipeline jobs locally, diagnoses each failure with a structured 4-part root cause analysis, and applies surgical fixes until every quality gate exits with code 0. ## Core Features & Use Cases - Local Runner Generation: Scans GitHub Actions, GitLab CI, Azure Pipelines, and other workflow files to generate a Python local runner (06-cicd-local-runner.py) with a concurrent worker pool, timeouts, and consolidated failure logs. - Autonomous Fix Loop: Runs a bounded self-loop that zeroes in on one failure at a time, writes a 4-part RCA (why, how, root cause, code fix) to .ai-memory/memory/issues/, applies a minimal fix, and re-verifies until exit code 0. - Remote Pipeline Diagnostics: Uses gitmap pipeline-ai status with adaptive ETA-based waiting to extract targeted failure lines from remote pipelines without polling loops. - Use Case: A screenshot shows a failing GitHub Actions workflow. The Skill extracts the pipeline name, updates the local runner's JOBS dictionary, reproduces the failure locally, fixes the offending code, and commits everything in one atomic commit. ## Quick Start Ask the AI to fix the failing CI/CD pipeline by running the local runner and resolving each error until all quality gates pass.
