NO-ISSUE: Release manager helper using Claude skills#6351
NO-ISSUE: Release manager helper using Claude skills#6351ggiguash wants to merge 9 commits intoopenshift:mainfrom
Conversation
openshift-ci-robot
added
the
jira/valid-reference
|
@ggiguash: This pull request explicitly references no jira issue. DetailsIn response to this: Instructions for interacting with me using PR comments are available here. If you have questions or suggestions related to my behavior, please file an issue against the openshift-eng/jira-lifecycle-plugin repository. |
|
Skipping CI for Draft Pull Request. |
openshift-ci
bot
added
the
do-not-merge/work-in-progress
|
Important Review skippedAuto reviews are limited based on label configuration. 🚫 Review skipped — only excluded labels are configured. (1)
Please check the settings in the CodeRabbit UI or the ⚙️ Run configurationConfiguration used: Repository YAML (base), Organization UI (inherited) Review profile: CHILL Plan: Pro Run ID: You can disable this status message by setting the Use the checkbox below for a quick retry:
WalkthroughAdds three CLI specification docs for CI analysis, two Bash scripts to query MicroShift Prow jobs (release and PR-focused), and updates .claude/settings.json to broaden permissions and register the new skills. Reports are produced under /tmp (text and self-contained HTML). Changes
Estimated code review effort🎯 3 (Moderate) | ⏱️ ~20 minutes ✨ Finishing Touches🧪 Generate unit tests (beta)
📝 Coding Plan
Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out. Comment |
|
[APPROVALNOTIFIER] This PR is APPROVED This pull-request has been approved by: ggiguash The full list of commands accepted by this bot can be found here. The pull request process is described here DetailsNeeds approval from an approver in each of these files:
Approvers can indicate their approval by writing |
openshift-ci
bot
added
the
approved
|
@coderabbitai perform full review |
|
✅ Actions performedFull review triggered. |
![coderabbitai[bot]](https://avatars.githubusercontent.com/in/347564?s=60&v=4){kind=small}
There was a problem hiding this comment.
Actionable comments posted: 4
🧹 Nitpick comments (3)
.claude/commands/analyze-ci-for-release-manager.md (1)
11-13: Add language identifiers to fenced code blocks.Static analysis flags missing language specifiers. Use
textorbashas appropriate.Also applies to: 38-40, 46-48, 190-198, 203-205, 208-210, 213-215
🤖 Prompt for AI Agents
Verify each finding against the current code and only fix it if needed. In @.claude/commands/analyze-ci-for-release-manager.md around lines 11 - 13, Several fenced code blocks in .claude/commands/analyze-ci-for-release-manager.md are missing language identifiers; update each triple-backtick block (e.g. the block containing `/analyze-ci-for-release-manager <release1,release2,...>`) to use an appropriate language tag such as `text` or `bash`, and apply the same change to the other reported blocks (lines around 38-40, 46-48, 190-198, 203-205, 208-210, 213-215) so all fenced code blocks include a language specifier..claude/commands/analyze-ci-for-release.md (1)
14-16: Add language identifiers to fenced code blocks.Static analysis flags these. Use
text,bash, or appropriate language.Also applies to: 50-52, 76-81, 131-182, 188-190, 199-201, 210-212, 236-239, 242-246, 249-252, 264-267, 270-273, 276-279
🤖 Prompt for AI Agents
Verify each finding against the current code and only fix it if needed. In @.claude/commands/analyze-ci-for-release.md around lines 14 - 16, Several fenced code blocks (for example the one showing "/analyze-ci-for-release <release-version> [--limit N]") are missing language identifiers; update every fenced block in this document to include an appropriate language tag (e.g., text, bash, json) so Markdown linters stop flagging them—search for the visible fenced snippets such as command usages, shell examples, and multi-line outputs (including the other blocks noted in the review) and add the correct language identifier after the opening ``` for each..claude/scripts/microshift-prow-jobs-for-release.sh (1)
30-61: Status-to-icon logic is duplicated.The same mapping exists in
status_to_icon()(unused) and inline in both jq queries. Consider consolidating.🤖 Prompt for AI Agents
Verify each finding against the current code and only fix it if needed. In @.claude/scripts/microshift-prow-jobs-for-release.sh around lines 30 - 61, Extract the duplicated status-to-icon logic into a single reusable jq function (e.g. def status_to_icon: if . == "success" then "✓" elif . == "failure" then "✗" elif . == "pending" then "⋯" else "?" end) and include that function at the top of both jq program strings used by query_status() and query_failed(), replacing the inline mapping expressions; also remove the unused shell-level status_to_icon() helper (or make it call through to the jq function if you prefer a shell wrapper) so the mapping is defined once and reused by both query_status() and query_failed().
🤖 Prompt for all review comments with AI agents
Verify each finding against the current code and only fix it if needed.
Inline comments:
In @.claude/commands/analyze-ci-for-release.md:
- Line 4: The description string in analyze-ci-for-release.md uses the wrong
skill name "analyse-ci-test-job"; update that description to use the exact skill
identifier "analyze-ci-test-job" so it matches the entry in
.claude/settings.json and avoids invocation failures—locate the description line
in analyze-ci-for-release.md and replace "analyse-ci-test-job" with
"analyze-ci-test-job".
In @.claude/scripts/microshift-prow-jobs-for-release.sh:
- Around line 13-16: The curl call that fetches "${PROW_URL}" can hang
indefinitely; update the curl invocation (the line using curl -s "${PROW_URL}"
before piping to jq) to include a timeout option (e.g. --max-time or
--connect-timeout and --max-time) so the script fails fast when the Prow server
is slow or unresponsive and still pipes output to jq for the existing filter
using ${release} and ${PROW_JOB}.
- Around line 93-95: The branch handling the --mode option assigns mode="${2}"
and then does shift 2 which will fail under set -u if no value is provided;
update the case arm for --mode to validate the next positional exists and is not
another flag before consuming it (for example check that "${2:-}" is non-empty
and does not start with '-' or ensure $# -ge 2), and if the check fails call the
usage/error path; otherwise set mode="$2" and shift 2 as before. Include
references to the --mode case, the mode variable, and the shift 2 operation so
reviewers can find and apply the guard.
In @.claude/settings.json:
- Around line 5-6: The setting "Bash(*)" in .claude/settings.json grants
unrestricted shell execution; restrict it by replacing "Bash(*)" with explicit,
minimal patterns for only the commands the tools need (for example
"Bash(curl:*)", "Bash(jq:*)", or similar command-specific patterns) and keep
"Write(//tmp/**)" if needed; update the Bash entry(s) to the smallest possible
whitelist to limit scope and remove any broad wildcard usage.
---
Nitpick comments:
In @.claude/commands/analyze-ci-for-release-manager.md:
- Around line 11-13: Several fenced code blocks in
.claude/commands/analyze-ci-for-release-manager.md are missing language
identifiers; update each triple-backtick block (e.g. the block containing
`/analyze-ci-for-release-manager <release1,release2,...>`) to use an appropriate
language tag such as `text` or `bash`, and apply the same change to the other
reported blocks (lines around 38-40, 46-48, 190-198, 203-205, 208-210, 213-215)
so all fenced code blocks include a language specifier.
In @.claude/commands/analyze-ci-for-release.md:
- Around line 14-16: Several fenced code blocks (for example the one showing
"/analyze-ci-for-release <release-version> [--limit N]") are missing language
identifiers; update every fenced block in this document to include an
appropriate language tag (e.g., text, bash, json) so Markdown linters stop
flagging them—search for the visible fenced snippets such as command usages,
shell examples, and multi-line outputs (including the other blocks noted in the
review) and add the correct language identifier after the opening ``` for each.
In @.claude/scripts/microshift-prow-jobs-for-release.sh:
- Around line 30-61: Extract the duplicated status-to-icon logic into a single
reusable jq function (e.g. def status_to_icon: if . == "success" then "✓" elif .
== "failure" then "✗" elif . == "pending" then "⋯" else "?" end) and include
that function at the top of both jq program strings used by query_status() and
query_failed(), replacing the inline mapping expressions; also remove the unused
shell-level status_to_icon() helper (or make it call through to the jq function
if you prefer a shell wrapper) so the mapping is defined once and reused by both
query_status() and query_failed().
ℹ️ Review info
⚙️ Run configuration
Configuration used: Repository YAML (base), Organization UI (inherited)
Review profile: CHILL
Plan: Pro
Run ID: 96f88164-5cb9-45af-821b-c07181a022a7
📒 Files selected for processing (4)
.claude/commands/analyze-ci-for-release-manager.md.claude/commands/analyze-ci-for-release.md.claude/scripts/microshift-prow-jobs-for-release.sh.claude/settings.json
|
/test verify |
|
✅ Actions performedFull review triggered. |
There was a problem hiding this comment.
Actionable comments posted: 5
🤖 Prompt for all review comments with AI agents
Verify each finding against the current code and only fix it if needed.
Inline comments:
In @.claude/commands/analyze-ci-for-release-manager.md:
- Around line 11-13: The markdown contains unlabeled fenced code blocks (for
example the fence with "/analyze-ci-for-release-manager
<release1,release2,...>") which triggers MD040; update each unlabeled
triple-backtick fence (including the other ranges mentioned) by adding an
appropriate language identifier such as bash, text, or html so the fences become
like ```bash or ```text; ensure you pick the label that best matches the snippet
content for consistent rendering and markdownlint compliance.
- Around line 163-164: The tab switching relies on an implicit global event;
update the inline handlers to pass the event explicitly (e.g., call
showTab(event, 'periodics') / showTab(event, 'pull-requests') instead of
showTab('...')) and change the showTab function signature to accept an event
parameter (e.g., function showTab(e, tabName)) and use e.target instead of
event.target; also update any other calls/usages of showTab (the other
occurrences around the tab-handling block) to the new two-argument form so they
no longer rely on the implicit global event.
In @.claude/commands/analyze-ci-for-release.md:
- Around line 22-23: Replace all occurrences of the mixed agent identifiers so
the document uses a single callable name "analyze-ci-test-job" everywhere
instead of "openshift-ci-test-job"; update every reference (e.g., the lines
showing "openshift-ci-test-job" and "analyze-ci-test-job" at the noted sections
and any headings, examples, or step lists) to the canonical
"analyze-ci-test-job" to ensure deterministic orchestration and consistent
invocation.
- Around line 14-16: Add explicit language identifiers to all fenced code blocks
flagged by markdownlint (MD040) so the blocks render and lint correctly; e.g.,
mark shell/command examples like "/analyze-ci-for-release <release-version>
[--limit N]" with a bash (or sh) tag and use text/plain or none for non-code
output blocks — update every fenced block mentioned in the review (including the
command example and the other flagged blocks) to include the appropriate
language tag.
In @.claude/settings.json:
- Around line 8-12: The entries Skill(analyze-ci-for-release-manager),
Skill(analyze-ci-for-release), Skill(analyze-ci-test-job),
Skill(analyze-ci-test-scenario), and Skill(analyze-sos-report) in the settings
are referencing command files (they exist under .claude/commands/) but not
actual skill/agent definitions, which will fail at runtime; fix by either (A)
adding matching agent/skill definitions named exactly
analyze-ci-for-release-manager, analyze-ci-for-release, analyze-ci-test-job,
analyze-ci-test-scenario, and analyze-sos-report under your agents/skills store
so Skill(...) points to real agents, or (B) remove these Skill(...) grants and
replace them with the appropriate slash-command invocations (or Command(...)
entries) that reference the existing command definitions (e.g., use the command
identifier for analyze-ci-for-release-manager etc.), ensuring the settings.json
entries match the actual artifact types present.
ℹ️ Review info
⚙️ Run configuration
Configuration used: Repository YAML (base), Organization UI (inherited)
Review profile: CHILL
Plan: Pro
Run ID: 70c7b89e-9758-4d69-b1c8-128c64cddd1b
📒 Files selected for processing (6)
.claude/commands/analyze-ci-for-pull-requests.md.claude/commands/analyze-ci-for-release-manager.md.claude/commands/analyze-ci-for-release.md.claude/scripts/microshift-prow-jobs-for-pull-requests.sh.claude/scripts/microshift-prow-jobs-for-release.sh.claude/settings.json
openshift-ci
bot
removed
the
do-not-merge/work-in-progress
|
/verified by @ggiguash |
openshift-ci-robot
added
the
verified
|
@ggiguash: This PR has been marked as verified by DetailsIn response to this:
Instructions for interacting with me using PR comments are available here. If you have questions or suggestions related to my behavior, please file an issue against the openshift-eng/jira-lifecycle-plugin repository. |
|
@ggiguash: all tests passed! Full PR test history. Your PR dashboard. DetailsInstructions for interacting with me using PR comments are available here. If you have questions or suggestions related to my behavior, please file an issue against the kubernetes-sigs/prow repository. I understand the commands that are listed here. |
2 participants
Overview
microshift-prow-jobs-for-release.shto extract the last failed jobs per given release to be analyzedmicroshift-prow-jobs-for-pull-request.shto extract the list of failed jobs for open pull requests, filtered by rebase PRs onlyanalyze-ci-for-releaseandanalyze-ci-for-pull-requestsskills use the existingopenshift-ci-analysisskill to run on the failed job URLs and produce text summariesanalyze-ci-for-release-managerskill allows analysis of multiple releases usinganalyze-ci-for-releaseandanalyze-ci-for-pull-requests, producing an HTML summary.Example of command execution
Mar 16 2026 Status of the example command
HTML summary
List of failed 4.18 jobs
List of open rebase PRs
$ .claude/scripts/microshift-prow-jobs-for-pull-requests.sh --filter 'NO-ISSUE: rebase-release' Fetching open PRs... Fetching job results... PR TITLE ✓ ✗ ⋯ JOBS https://github.com/openshift/microshift/pull/6357 NO-ISSUE: rebase-release-4.21-4.21.0-0.nightly-... 7 0 6 13 https://github.com/openshift/microshift/pull/6356 NO-ISSUE: rebase-release-4.18-4.18.0-0.nightly-... 6 0 6 12 https://github.com/openshift/microshift/pull/6355 NO-ISSUE: rebase-release-4.19-4.19.0-0.nightly-... 7 0 6 13 https://github.com/openshift/microshift/pull/6354 NO-ISSUE: rebase-release-4.20-4.20.0-0.nightly-... 7 0 6 13 https://github.com/openshift/microshift/pull/6353 NO-ISSUE: rebase-release-4.17-4.17.0-0.nightly-... 6 0 13 19