NO-JIRA: claude: Turn pr-review skill into code-reviewer agent

[tchap]

There is no reason to fill your context window with code review stuff, better to split it into a subagent.

[openshift-ci-robot]

@tchap: This pull request explicitly references no jira issue.

Details

In response to this:

There is no reason to fill your context window with code review stuff, better to split it into a subagent.

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.

[coderabbitai]

No actionable comments were generated in the recent review. 🎉

ℹ️ Recent review info

⚙️ Run configuration

Configuration used: Repository YAML (base), Organization UI (inherited)

Review profile: CHILL

Plan: Pro

Run ID: f5c192ba-adc1-4a8f-9ccd-685570d694d4

📥 Commits

Reviewing files that changed from the base of the PR and between 7b89009 and b7c53a3.

📒 Files selected for processing (4)

• .claude/agents/code-reviewer.md
• .claude/commands/pr-review.md
• .claude/settings.json
• .claude/skills/pr-review/SKILL.md

💤 Files with no reviewable changes (2)

• .claude/commands/pr-review.md
• .claude/skills/pr-review/SKILL.md

✅ Files skipped from review due to trivial changes (2)

• .claude/settings.json
• .claude/agents/code-reviewer.md

---

Walkthrough

Adds a new Claude "code-reviewer" agent configuration and updates agent permissions; also removes an existing pr-review command and its associated skill file. The agent runs automated checks (build/test/verify/tidy and git diff) and performs oc/kubectl-aware code reviews.

Changes

Cohort / File(s)	Summary
Agent Configuration .claude/agents/code-reviewer.md	Added new Claude agent defining a code-reviewer role that runs go mod tidy -diff, make oc, make verify, make test, and performs git-diff-based oc/kubectl-aware code review with structured findings.
Permissions Settings .claude/settings.json	Added top-level permissions.allow entries permitting execution of Bash(make oc), Bash(make verify), Bash(make test), Bash(go mod tidy -diff), and Bash(git diff:*).
Removed Command .claude/commands/pr-review.md	Removed the pr-review command metadata/documentation (frontmatter, synopsis, usage and argument specification).
Removed Skill .claude/skills/pr-review/SKILL.md	Deleted the PR Review skill document that described the previous end-to-end review workflow and guidance (tidy/build/verify/test, code-quality checks, and summary format).

Estimated code review effort

🎯 2 (Simple) | ⏱️ ~10 minutes

✨ Finishing Touches

🧪 Generate unit tests (beta)

• Create PR with unit tests

---

_{Comment @coderabbitai help to get the list of available commands and usage tips.}

[coderabbitai]

Actionable comments posted: 3

🧹 Nitpick comments (1)

.claude/agents/code-reviewer.md (1)

20-27: “Run in parallel” is ambiguous and may reduce signal quality.

For these heavyweight targets, forcing parallel execution can increase contention and make failures harder to interpret. Consider stating an explicit order (or parallel groups) instead of blanket parallelism.

🤖 Prompt for AI Agents

Verify each finding against the current code and only fix it if needed.

In @.claude/agents/code-reviewer.md around lines 20 - 27, The instruction "Run
these commands in parallel" is ambiguous and risky for heavyweight targets;
change the guidance to an explicit order or defined parallel groups instead of
blanket parallelism—for example, run "go mod tidy -diff" first, then "make oc",
and finally run "make verify" and "make test" sequentially (or state them as a
controlled parallel group if needed), updating the text that lists "go mod tidy
-diff", "make oc", "make verify", and "make test" accordingly to reflect the
chosen order/grouping.

🤖 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/agents/code-reviewer.md:
- Around line 60-64: The summary currently mandates a standalone "Linting
status" but no explicit lint step exists; update the reporting template or CI to
be consistent by either adding a lint job (run the linter and report its
pass/fail) or renaming the field to "Verify/Lint status (from `make verify`)"
and populate it from the existing `make verify` result; ensure the structured
summary output (Build status, Test results, Linting/Verify status, Code quality
observations) uses that single, unambiguous field so consumers know whether
linting was actually executed.
- Around line 29-31: The current known-issue line that broadly ignores failures
under the package string `github.com/openshift/oc/pkg/cli` and the phrase
"kubeconfig error" is too wide; update that sentence to target the specific
failing test and/or error signature instead (e.g., reference the failing test
function name and include a precise error substring or regex like
"kubeconfig.*no such file or directory" or the exact test identifier) so only
that test failure is exempted; edit the line that currently reads "Known Issue:
Test failure in `github.com/openshift/oc/pkg/cli` (kubeconfig error) can be
ignored" to instead name the exact test and error pattern to match.

In @.claude/settings.json:
- Around line 2-8: The permissions allowlist under the "permissions" -> "allow"
array is missing the required Git permission; add "Bash(git diff)" to the allow
list (alongside existing entries like "Bash(make oc)", "Bash(make verify)",
"Bash(make test)", "Bash(go mod tidy -diff)") so the code-reviewer agent can run
its core git diff step referenced in agents/code-reviewer.md; update the "allow"
array in .claude/settings.json accordingly.

---

Nitpick comments:
In @.claude/agents/code-reviewer.md:
- Around line 20-27: The instruction "Run these commands in parallel" is
ambiguous and risky for heavyweight targets; change the guidance to an explicit
order or defined parallel groups instead of blanket parallelism—for example, run
"go mod tidy -diff" first, then "make oc", and finally run "make verify" and
"make test" sequentially (or state them as a controlled parallel group if
needed), updating the text that lists "go mod tidy -diff", "make oc", "make
verify", and "make test" accordingly to reflect the chosen order/grouping.

🪄 Autofix (Beta)

Fix all unresolved CodeRabbit comments on this PR:

• Push a commit to this branch (recommended)
• Create a new PR with the fixes

---

ℹ️ Review info

⚙️ Run configuration

Configuration used: Repository YAML (base), Organization UI (inherited)

Review profile: CHILL

Plan: Pro

Run ID: 5b018651-c1ca-4434-9881-5b148f374481

📥 Commits

Reviewing files that changed from the base of the PR and between c268e49 and 7b89009.

📒 Files selected for processing (2)

• .claude/agents/code-reviewer.md
• .claude/settings.json

[tchap]

I will continue with CodeRabbit review tomorrow.

/retitle WIP: NO-JIRA: claude: Turn pr-review skill into code-reviewer agent

[ardaguclu]

One comment. Also would it make sense updating the README about how this can be used?.

[tchap]

/retitle NO-JIRA: claude: Turn pr-review skill into code-reviewer agent

[ardaguclu]

This PR looks good to me. Maybe it is better to check the reviews from CodeRabbit. Ping me once this is ready to tag.

[tchap]

@coderabbitai review

[coderabbitai]

✅ Actions performed

Review triggered.

Note: CodeRabbit is an incremental review system and does not re-review already reviewed commits. This command is applicable only when automatic reviews are paused.

[ardaguclu]

That looks great, we are one step closer to agentic sdlc :). Thank you
/lgtm
/approve

[openshift-ci]

[APPROVALNOTIFIER] This PR is APPROVED

This pull-request has been approved by: ardaguclu, tchap

The full list of commands accepted by this bot can be found here.

The pull request process is described here

Details

Needs approval from an approver in each of these files:

• OWNERS [ardaguclu]

Approvers can indicate their approval by writing /approve in a comment
Approvers can cancel approval by writing /approve cancel in a comment

[tchap]

/verified by @tchap

[openshift-ci-robot]

@tchap: This PR has been marked as verified by @tchap.

Details

In response to this:

/verified by @tchap

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.

[openshift-ci-robot]

/retest-required

Remaining retests: 0 against base HEAD bb4c3de and 2 for PR HEAD b7c53a3 in total

[tchap]

/retest

[openshift-ci]

@tchap: all tests passed!

Full PR test history. Your PR dashboard.

Details

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 kubernetes-sigs/prow repository. I understand the commands that are listed here.