Triage Ci Flake

The triage-ci-flake skill enables AI agents to automatically detect and analyze flaky tests within Continuous Integration pipelines. It helps maintain code quality by identifying unstable test cases that cause false failures, allowing teams to focus on genuine bugs rather than noise.

When to use it

• Automate Test Stability: Run during nightly builds to catch intermittent failures before they accumulate in the main branch.
• Reduce Noise: Filter out non-critical CI alerts so engineers can prioritize actual production issues.
• Improve Pipeline Reliability: Identify flaky tests that block deployments or waste developer time on false positives.
• Maintain Historical Data: Track patterns of flakiness over time to improve test suite robustness.

Key capabilities

• Detects flaky tests in CI environments
• Analyzes test execution logs for instability patterns
• Provides actionable insights to fix or quarantine unstable tests

Example prompts

• "Analyze the last 50 CI runs and identify any tests that failed intermittently."
• "List all flaky tests detected in the current build pipeline along with their failure frequency."
• "Generate a report of unstable test cases that need immediate attention or quarantine."

Tips & gotchas

Ensure your CI logs are accessible and contain sufficient detail for the agent to distinguish between random failures and actual bugs. This skill works best when integrated early in the development cycle to prevent flakiness from becoming systemic.