Agent Skills: General instructions

Analyze test errors

UncategorizedID: openshift/hypershift/e2e-analyze

Repository

openshiftLicense: Apache-2.0
541567

Install this agent skill to your local

pnpm dlx add-skill https://github.com/openshift/hypershift/tree/HEAD/.claude/skills/e2e-analyze

Skill Files

Browse the full folder contents for e2e-analyze.

Download Skill

Loading file tree…

.claude/skills/e2e-analyze/SKILL.md

Skill Metadata

Name
e2e-analyze
Description
"Analyze test errors"

General instructions

  • Validate that "$3" is an existing directory; fail if it is not. Do NOT create the directory.
  • Extract {BUILD_NUMBER} from "$1" by matching a 10–20 digit sequence; fail if none is found.
  • Use the extracted number as {BUILD_NUMBER}.
  • Use curl to download the file build-log.txt under "$1". Only allow HTTPS URLs (reject HTTP).
  • When fetching, use: curl -fsSL --max-time 20 --retry 3 --retry-connrefused --proto '=https' --max-filesize 100M "${1}/build-log.txt".
  • The build-log.txt file contains e2e failures, store the artifacts related to failure "$2" under directory specified by "$3" to determine a possible root cause for the failure.
  • Use the "gcloud storage" command to fetch given artifacts under "$3" and make sure to use {BUILD_NUMBER} in URLs.
  • Provide evidence for the failure.
  • Try to find additional evidence. For example, in logs and events.
  • Do not delete downloaded artifacts during the cleanup phase.

Output format

  • The output should be formatted as:
    Error: {Error message here}
    Summary: {Failure analysis here}
    Evidence: {Evidence here}
    Additional evidence: {Additional evidence here}