Can I filter the results to only see high-certainty matches?

Yes, you can use the --min-confidence argument (e.g., --min-confidence 5) to filter out lower-confidence suggestions and focus on the most likely matches.

Do I need an API key or authentication to use this skill?

No, this skill uses the production Sippy URL which does not require authentication for querying related regression matches.

What output formats are available?

The skill supports 'json' for programmatic use and 'summary' for a human-readable report of the findings.

How does the skill identify 'related' regressions?

It uses two primary strategies: comparing test names using edit distance to find similar titles and matching regressions that share the exact same last failure timestamp, indicating they failed in the same job runs.

What is a 'confidence level' in the context of this skill?

Confidence levels range from 1 to 10. A score of 10 indicates a high-confidence match (exact test ID), 5 indicates a medium-confidence match (similarly named tests), and 2 indicates a low-confidence match (shared failure timestamp).

Fetch Related Triages

Name: Fetch Related Triages
Author: openshift-eng

byopenshift-eng

•

部署与 DevOps

Identifies existing triage records and related untriaged regressions to streamline CI/CD failure analysis.

This skill connects to the Sippy API to analyze Component Readiness regressions, helping developers find existing JIRA triages or shared failure patterns. By matching similarly named tests and shared failure timestamps, it enables users to avoid filing duplicate bugs, identify common root causes across multiple failures, and group related regressions for more efficient resolution within the OpenShift CI ecosystem. It is an essential tool for developers and QE engineers who need to quickly determine if a CI failure is a known issue or a new regression requiring investigation.

主要功能

01Queries Sippy API for existing triage records and JIRA associations

02Calculates test name similarity using edit distance to find related failures

03Provides both machine-readable JSON and human-friendly summary outputs

0460 GitHub stars

05Identifies untriaged regressions with matching failure timestamps

06Supports filtering by confidence levels (1-10) for high-precision matching

使用场景

01Consolidating multiple related CI regressions into a single triage effort

02Determining if a regression is a localized issue or part of a broader infrastructure failure

03Searching for existing JIRA bugs that might already cover a new CI failure

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add openshift-eng/ai-helpers fetch-related-triages

For use in Claude.ai and ChatGPT

Download Skill