From 0c58750162e767a0a2edefd0da29731878adfed2 Mon Sep 17 00:00:00 2001 From: Stephen Benjamin Date: Thu, 19 Mar 2026 06:15:00 -0400 Subject: [PATCH] feat(ci): treat openshift/release PRs as revert candidates in analyze-payload When a subagent traces a failure root cause to a PR outside the payload (e.g., an openshift/release step registry change), include it as a candidate scored the same way as payload PRs rather than dismissing it as an infrastructure issue. Co-Authored-By: Claude Opus 4.6 --- .claude-plugin/marketplace.json | 2 +- docs/data.json | 2 +- plugins/ci/.claude-plugin/plugin.json | 2 +- plugins/ci/skills/analyze-payload/SKILL.md | 4 +++- plugins/ci/skills/prow-job-analyze-test-failure/SKILL.md | 1 + 5 files changed, 7 insertions(+), 4 deletions(-) diff --git a/.claude-plugin/marketplace.json b/.claude-plugin/marketplace.json index 23659d6da..0f7d02d4e 100644 --- a/.claude-plugin/marketplace.json +++ b/.claude-plugin/marketplace.json @@ -26,7 +26,7 @@ "name": "ci", "source": "./plugins/ci", "description": "A plugin to work with OpenShift CI and analyze Prow job results", - "version": "0.0.32" + "version": "0.0.33" }, { "name": "teams", diff --git a/docs/data.json b/docs/data.json index 10430002a..b052edbc0 100644 --- a/docs/data.json +++ b/docs/data.json @@ -563,7 +563,7 @@ "name": "Trigger Payload Job" } ], - "version": "0.0.32" + "version": "0.0.33" }, { "commands": [ diff --git a/plugins/ci/.claude-plugin/plugin.json b/plugins/ci/.claude-plugin/plugin.json index e86f86891..84be2a6e4 100644 --- a/plugins/ci/.claude-plugin/plugin.json +++ b/plugins/ci/.claude-plugin/plugin.json @@ -1,7 +1,7 @@ { "name": "ci", "description": "Tools for working with OpenShift CI and analyzing Prow job results", - "version": "0.0.32", + "version": "0.0.33", "author": { "name": "openshift" } diff --git a/plugins/ci/skills/analyze-payload/SKILL.md b/plugins/ci/skills/analyze-payload/SKILL.md index af89439e8..0527316ba 100644 --- a/plugins/ci/skills/analyze-payload/SKILL.md +++ b/plugins/ci/skills/analyze-payload/SKILL.md @@ -197,7 +197,9 @@ Wait for all subagents to complete and collect their analysis results. For each #### 6.1: Correlate Failures with Candidate PRs -For each failed job, cross-reference the failure analysis from the subagent with the candidate PRs from the originating payload. Score each (failed job, candidate PR) pair using the following weighted rubric: +For each failed job, cross-reference the failure analysis from the subagent with the candidate PRs from the originating payload. Additionally, if a subagent traced the root cause to a PR outside the payload (e.g., an `openshift/release` PR that modified a CI step registry script), include that PR as a candidate — it is a regression like any other and should be scored and treated the same way as payload PRs. + +Score each (failed job, candidate PR) pair using the following weighted rubric: | Signal | Weight | Criteria | |--------|--------|----------| diff --git a/plugins/ci/skills/prow-job-analyze-test-failure/SKILL.md b/plugins/ci/skills/prow-job-analyze-test-failure/SKILL.md index fb8676b1b..79a21de0c 100644 --- a/plugins/ci/skills/prow-job-analyze-test-failure/SKILL.md +++ b/plugins/ci/skills/prow-job-analyze-test-failure/SKILL.md @@ -336,6 +336,7 @@ The CI system may attach **symptom labels** to job runs — machine-detected pat - Store artifacts from Prow CI job (json/yaml files) related to the failure under `.work/prow-job-analyze-test-failure/{build_id}/tmp` - Store logs under `.work/prow-job-analyze-test-failure/{build_id}/logs/` - Collect evidence from logs and events and other json/yaml files +- **Check for CI step script errors**: If the error is a scripting issue in a CI step (e.g., unbound variable, syntax error, missing command, bad exit code from a shell script) rather than a product bug, check for recent commits to that step's script in the `openshift/release` repository. This is part of evidence gathering — identifying the responsible PR early informs the rest of the analysis. Include the responsible PR in your evidence. ### Step 4.4b: Investigate crash-looping or failing containers