docs(workflow): require PRs to mirror their closing issue's triage labels (closes #237) by cristim · Pull Request #238 · LeanerCloud/cloud-commitments-cli

cristim · 2026-05-03T11:47:10Z

Summary

Closes docs(workflow): require PRs to mirror their closing issue's triage labels #237: codifies three mandatory workflow rules that were present in the system-prompt CLAUDE.md but absent from the feat/multicloud-web-frontend base branch.
Post-push CI watcher — arm a background watcher after every git push so CI failures surface within the same turn.
CodeRabbit loop — iterate CR passes until zero Actionable items remain before declaring a PR done.
PR labeling — mirror the closing issue's triage labels (priority/*, severity/*, urgency/*, impact/*, effort/*, type/*, triaged) onto every new PR via gh pr edit --add-label, before pinging CodeRabbit. Unlabeled PRs are invisible to priority-ordered is:open label:priority/p* queries.

The silent failure that motivated this PR: 10 PRs shipped in two recent batches, all merged-eligible after CR sign-off, none labeled — making them invisible to triage ordering.

Dogfooding

This PR itself is labeled with the issue's triage labels per the new rule (type/docs, priority/p3, severity/low, urgency/eventually, impact/internal, effort/xs, triaged).

Test plan

CLAUDE.md contains ## PR labeling — mirror closing-issue labels (MANDATORY) between ## CodeRabbit loop and ## Security Rules
Section includes the gh invocation with jq filter for all label classes
Multi-close handling (highest priority/severity, union the rest) is documented
No-triaged-lie rule is present
Closing checkpoint paragraph matches the CI-watcher pattern
Post-push CI watcher and CodeRabbit loop sections also present (they were absent from base branch)

Summary by CodeRabbit

Documentation
- Added a mandatory post-push CI monitoring procedure that identifies queued/in-progress runs, reports outcomes, surfaces failure details, and triggers corrective fix-pushes when needed.
- Introduced a mandatory iterative review loop that repeats until all actionable items are resolved and nitpicks are either fixed or explicitly justified, following defined batching rules.
- Established PR labeling requirements to mirror associated issue triage labels, including multi-issue handling and clear verification steps for label application.

coderabbitai · 2026-05-03T11:47:15Z

No actionable comments were generated in the recent review. 🎉

ℹ️ Recent review info

⚙️ Run configuration

Configuration used: defaults

Review profile: CHILL

Plan: Pro

Run ID: 3297c8fe-feb2-4cd3-b310-6a961906cf06

📥 Commits

Reviewing files that changed from the base of the PR and between 64fc92a and 30a5a6f.

📒 Files selected for processing (1)

CLAUDE.md

🚧 Files skipped from review as they are similar to previous changes (1)

CLAUDE.md

📝 Walkthrough

Walkthrough

Documentation adds mandatory post-push CI watching, CodeRabbit review iteration, and closing-issue label mirroring procedures to CLAUDE.md.

Changes

Workflow Procedure Documentation

Layer / File(s)	Summary
Post-Push CI Watcher `CLAUDE.md`	Lists push-triggered GitHub Actions runs, launches background watchers for `queued`/`in_progress` runs, reports failed step digests, and performs automated fix-pushes.
CodeRabbit Review Loop `CLAUDE.md`	Defines review triggering, polling, finding triage, fix-push handling, conflict recovery, and repeat-until-clean exit criteria.
PR Labeling — Mirror Closing Issue `CLAUDE.md`	Defines closing-issue label mirroring during `gh pr create`, verification, missing-label handling, multi-issue rules, and final confirmation.

Estimated code review effort: 2 (Simple) | ~10 minutes

🚥 Pre-merge checks | ✅ 5

✅ Passed checks (5 passed)

Check name	Status	Explanation
Description Check	✅ Passed	Check skipped - CodeRabbit’s high-level summary is enabled.
Title check	✅ Passed	The title clearly states the main workflow change: mirroring closing issue triage labels onto PRs.
Linked Issues check	✅ Passed	The PR adds the mandatory labeling section, multi-close handling, and cross-references required by `#237`.
Out of Scope Changes check	✅ Passed	The CI watcher and CodeRabbit loop are referenced acceptance criteria, so they are in scope rather than extraneous.
Docstring Coverage	✅ Passed	No functions found in the changed files to evaluate docstring coverage. Skipping docstring coverage check.

✨ Finishing Touches

🧪 Generate unit tests (beta)

Create PR with unit tests
Commit unit tests in branch docs/issue-237-codify-pr-labeling

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

cristim · 2026-05-03T11:47:34Z

@coderabbitai review

coderabbitai · 2026-05-03T11:47:39Z

✅ Actions performed

Review triggered.

coderabbitai

Actionable comments posted: 1

🧹 Nitpick comments (1)

CLAUDE.md (1)

150-153: ⚡ Quick win

Avoid manual PR_NUM entry in a mandatory path.

Using a hand-copied PR number is brittle and easy to misapply to the wrong PR. Prefer deriving it directly from the current branch context.

Safer PR number resolution

-PR_NUM=<the number gh just printed>
+PR_NUM=$(gh pr view --repo LeanerCloud/CUDly --json number --jq '.number')
 ISSUE_NUM=<the issue this PR closes>

🤖 Prompt for AI Agents

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

In `@CLAUDE.md` around lines 150 - 153, Replace the brittle manual PR_NUM
assignment after `gh pr create` with automatic resolution: capture the created
PR number programmatically (e.g., parse `gh pr create` output or call `gh pr
view` for the current branch) and set PR_NUM from that value; reference the
variables PR_NUM and ISSUE_NUM and the command `gh pr create` in your changes so
the script derives PR_NUM from the current branch/PR context rather than
requiring manual entry.

🤖 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.md`:
- Around line 154-160: The script currently unconditionally runs gh pr edit
"$PR_NUM" --add-label "$LABELS", which fails when LABELS is empty; change it to
guard the call by checking the LABELS variable first (the LABELS assignment and
the gh pr edit "$PR_NUM" --add-label "$LABELS" invocation), e.g., only call gh
pr edit with --add-label when LABELS is non-empty and otherwise surface the
triage gap deterministically (for example add a specific fallback label like
"triage-needed" or post a comment) so the workflow does not break on empty label
payloads.

---

Nitpick comments:
In `@CLAUDE.md`:
- Around line 150-153: Replace the brittle manual PR_NUM assignment after `gh pr
create` with automatic resolution: capture the created PR number
programmatically (e.g., parse `gh pr create` output or call `gh pr view` for the
current branch) and set PR_NUM from that value; reference the variables PR_NUM
and ISSUE_NUM and the command `gh pr create` in your changes so the script
derives PR_NUM from the current branch/PR context rather than requiring manual
entry.

🪄 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: defaults

Review profile: CHILL

Plan: Pro

Run ID: 00bc1995-90e3-4a99-982a-b2b93e9aaacd

📥 Commits

Reviewing files that changed from the base of the PR and between c84fd02 and 7dbe911.

📒 Files selected for processing (1)

CLAUDE.md

cristim · 2026-05-03T11:50:12Z

Added a follow-up commit (188012004) extending the CR-loop section with explicit subagent-delegation guidance — exit criteria for fork prompts, not just human workflows. Surfaced by the catch-up CR-loop pass on PRs #227–#236 (none of those forks ran the CR loop because my prompts stopped at the @coderabbitai review ping). @coderabbitai review

coderabbitai · 2026-05-03T11:50:23Z

Tip

For best results, initiate chat on the files or code changes.

@cristim: Understood — triggering a review pass to cover commit 188012004 and the subagent-delegation exit-criteria additions to the CR-loop section.

[review]

…nd manual PR_NUM Address CodeRabbit review on PR #238: 1. Actionable (Major): `gh pr edit --add-label ""` fails on empty payload, silently breaking the MANDATORY label-mirror flow when the closing issue has no priority/severity/urgency/impact/effort/type labels. Guard the call with `if [ -n "$LABELS" ]` and emit a clear WARN otherwise that directs the operator to triage the issue first and surface the gap. 2. Nitpick (Quick win): `PR_NUM=<the number gh just printed>` is brittle and easy to misapply to the wrong PR. Derive it from the current branch context via `gh pr view --json number --jq '.number'` instead, so the snippet works without any hand-editing after `gh pr create`. Both changes are in the prescriptive CLAUDE.md snippet only. The labels flow itself is unchanged behaviorally for the happy path (issue has the expected triage labels); the new branches harden the unhappy paths. Refs #237.

cristim · 2026-05-19T15:13:19Z

CodeRabbit pass 1 — both findings addressed

Pushed commit 64fc92a addressing both items from the 2026-05-03 review:

Actionable (Major) — empty `$LABELS` guard

Fixed. gh pr edit --add-label "" does fail (silently breaks the MANDATORY flow), so I added an if [ -n "$LABELS" ] guard with a clear WARN fallback that names the missing classes and tells the operator to triage the issue first. Matches your proposed hardening with slightly more informative output (the WARN message explicitly references the issue number and the expected label classes, so the operator knows what's missing without re-reading CLAUDE.md).

Nitpick (Quick win) — manual `PR_NUM` entry

Fixed. Replaced PR_NUM=<the number gh just printed> with PR_NUM=$(gh pr view --repo LeanerCloud/CUDly --json number --jq '.number') exactly as you suggested. Removes the hand-copy step entirely — the snippet now works without any operator edits after gh pr create (since gh pr view defaults to the current branch).

ISSUE_NUM is intentionally kept as a placeholder — there's no reliable way to auto-derive the closing-issue number from the branch alone, and the PR body's Closes #N line isn't structured enough to parse safely. The operator still has to fill that in once.

CI green on the previous head; watchers armed for 64fc92a. No behavioural changes to the happy path.

@coderabbitai full review

coderabbitai · 2026-05-19T15:15:08Z

Tip

For best results, initiate chat on the files or code changes.

🧩 Analysis chain

🏁 Script executed:

#!/bin/bash
# Fetch the CLAUDE.md from the PR branch
git log --oneline -8
echo "---"
cat CLAUDE.md

Length of output: 19004

@cristim: Full review complete on commits 7dbe911 + 1880120 + 64fc92a. Both prior findings are cleanly resolved and the three new CLAUDE.md sections satisfy all acceptance criteria from #237.

Review summary

✅ Empty `$LABELS` guard (was Actionable/Major)

The if [ -n "$LABELS" ] guard is correct. When the closing issue has no labels in the selected classes, jq's join(",") on an empty array produces "", which fails [ -n "" ], and the WARN path fires with a message that names the issue number and the expected label classes. No silent breakage of the MANDATORY flow.

✅ `PR_NUM` auto-derivation (was Nitpick)

PR_NUM=$(gh pr view --repo LeanerCloud/CUDly --json number --jq '.number') is correct. gh pr view defaults to the current branch's PR (which exists by the time this runs, since it's post–gh pr create). The hand-copy step is gone.

✅ `ISSUE_NUM` as intentional placeholder

Keeping ISSUE_NUM=<the issue this PR closes> is the right call. There is no reliable machine-parseable link between a branch name and its closing issue number, and parsing the PR body's Closes #N`` line is fragile at creation time. One manual fill-in is a reasonable trade-off.

✅ `jq` filter correctness

[.labels[].name | select(test("^(priority|severity|urgency|impact|effort|type)/")) ]
+ (if [.labels[].name] | any(. == "triaged") then ["triaged"] else [] end)
| join(",")

any(. == "triaged") desugars to any(.[]; . == "triaged") — valid jq, correctly detects the triaged label.
gh issue view --json labels always returns a labels array (never null), so no null-dereference risk.
join(",") on the concatenated array produces a gh pr edit --add-label-compatible comma-separated string.

✅ Acceptance criteria from `#237`

Criterion	Status
"PR labeling — mirror closing-issue labels (MANDATORY)" section	✅
Exact `gh` invocation with `jq` filter for all label classes	✅
Multi-close handling (highest priority/severity, union rest)	✅
No-triaged-lie rule	✅
"Before declaring done" checkpoint mirrors CI-watcher pattern	✅
Post-push CI watcher section present	✅
CodeRabbit loop section present	✅
Subagent-delegation exit criteria in CR-loop	✅

No actionable items remain. The PR is ready to merge.

[approve]

cristim · 2026-05-20T20:54:10Z

@coderabbitai review

coderabbitai · 2026-05-20T20:54:15Z

✅ Actions performed

Review triggered.

cristim · 2026-05-20T23:23:24Z

@coderabbitai resolve

coderabbitai · 2026-05-20T23:23:31Z

✅ Actions performed

Comments resolved. Approval is disabled; enable reviews.request_changes_workflow to allow explicit top-level @coderabbitai resolve or @coderabbitai approve commands.

cristim · 2026-05-21T12:26:46Z

@coderabbitai review

coderabbitai · 2026-05-21T12:26:52Z

✅ Actions performed

Review triggered.

cristim · 2026-05-21T12:55:44Z

@coderabbitai review

coderabbitai · 2026-05-21T12:55:50Z

✅ Actions performed

Review triggered.

cristim · 2026-05-22T10:28:13Z

@coderabbitai review

coderabbitai · 2026-05-22T10:28:19Z

✅ Actions performed

Review triggered.

cristim · 2026-05-22T13:08:52Z

CR finding verification - PR_NUM auto-derivation

Both findings from the CodeRabbit review (2026-05-03) against commit 7dbe911 are already addressed in commit 64fc92a:

Nitpick (lines 150-153): brittle manual PR_NUM
Fixed by replacing PR_NUM=<the number gh just printed> with:

PR_NUM=$(gh pr view --repo LeanerCloud/CUDly --json number --jq '.number')

This derives the PR number automatically from the current branch context.

Major (lines 165-180): unconditional gh pr edit --add-label on empty LABELS
Fixed by guarding with if [ -n "$LABELS" ] and printing a deterministic warning when no matching triage labels exist on the closing issue.

No new commits are needed. Current HEAD is 64fc92a on branch docs/issue-237-codify-pr-labeling.

@coderabbitai review

coderabbitai · 2026-05-22T13:09:13Z

@cristim: Confirmed — your verification summary matches the conclusions from the full review pass on 64fc92a:

Nitpick (lines 150–153): PR_NUM auto-derivation via gh pr view --repo LeanerCloud/CUDly --json number --jq '.number' correctly resolves the hand-copy fragility. ✅
Major (lines 165–180): The if [ -n "$LABELS" ] guard with the informative WARN path (naming the issue number and expected label classes) correctly prevents silent failure on an empty label set. ✅

Both fixes were already validated in the prior review, all acceptance criteria from #237 are satisfied, and no actionable items remain. Triggering a fresh incremental pass to cover the current HEAD.

✅ Actions performed

Review triggered.

cristim · 2026-05-22T23:00:29Z

@coderabbitai review

coderabbitai · 2026-05-22T23:00:34Z

✅ Actions performed

Review triggered.

cristim · 2026-06-04T14:04:52Z

@coderabbitai review

coderabbitai · 2026-06-04T14:04:58Z

Rate Limit Exceeded

@cristim have exceeded the limit for the number of chat messages per hour. Please wait 38 minutes and 37 seconds before sending another message.

cristim · 2026-06-06T06:39:24Z

@coderabbitai review

coderabbitai · 2026-06-06T06:39:29Z

✅ Action performed

Review finished.

cristim · 2026-06-07T23:02:45Z

@coderabbitai full review

coderabbitai · 2026-06-07T23:02:51Z

✅ Action performed

Full review finished.

…nd manual PR_NUM Address CodeRabbit review on PR #238: 1. Actionable (Major): `gh pr edit --add-label ""` fails on empty payload, silently breaking the MANDATORY label-mirror flow when the closing issue has no priority/severity/urgency/impact/effort/type labels. Guard the call with `if [ -n "$LABELS" ]` and emit a clear WARN otherwise that directs the operator to triage the issue first and surface the gap. 2. Nitpick (Quick win): `PR_NUM=<the number gh just printed>` is brittle and easy to misapply to the wrong PR. Derive it from the current branch context via `gh pr view --json number --jq '.number'` instead, so the snippet works without any hand-editing after `gh pr create`. Both changes are in the prescriptive CLAUDE.md snippet only. The labels flow itself is unchanged behaviorally for the happy path (issue has the expected triage labels); the new branches harden the unhappy paths. Refs #237.

cristim · 2026-06-19T15:43:20Z

@coderabbitai review

coderabbitai · 2026-06-19T15:43:27Z

✅ Action performed

Review finished.

cristim · 2026-07-09T21:20:38Z

@coderabbitai review

coderabbitai · 2026-07-09T21:20:45Z

✅ Action performed

Review finished.

…nd manual PR_NUM Address CodeRabbit review on PR #238: 1. Actionable (Major): `gh pr edit --add-label ""` fails on empty payload, silently breaking the MANDATORY label-mirror flow when the closing issue has no priority/severity/urgency/impact/effort/type labels. Guard the call with `if [ -n "$LABELS" ]` and emit a clear WARN otherwise that directs the operator to triage the issue first and surface the gap. 2. Nitpick (Quick win): `PR_NUM=<the number gh just printed>` is brittle and easy to misapply to the wrong PR. Derive it from the current branch context via `gh pr view --json number --jq '.number'` instead, so the snippet works without any hand-editing after `gh pr create`. Both changes are in the prescriptive CLAUDE.md snippet only. The labels flow itself is unchanged behaviorally for the happy path (issue has the expected triage labels); the new branches harden the unhappy paths. Refs #237.

…oses #237 Add project-specific guidance to CLAUDE.md for the post-push CI watcher (mandatory for all commits on CUDly PRs due to 9-15 min pre-commit job runtime) and label-mirroring rule (PRs must carry the same priority/severity/ urgency/impact/effort/type labels as their closing issues). Resolves conflict by including main's "Known Issues" section alongside the new sections added by this PR.

cristim · 2026-07-19T20:11:45Z

Rebased onto origin/main, all CI checks passed, squash-merged.

…oses #237 (#238) Add project-specific guidance to CLAUDE.md for the post-push CI watcher (mandatory for all commits on CUDly PRs due to 9-15 min pre-commit job runtime) and label-mirroring rule (PRs must carry the same priority/severity/ urgency/impact/effort/type labels as their closing issues). Resolves conflict by including main's "Known Issues" section alongside the new sections added by this PR.

cristim added type/docs Documentation priority/p3 Polish / idea / may never ship severity/low Minor harm urgency/eventually No deadline impact/internal Team-internal only effort/xs Trivial / one-liner triaged Item has been triaged labels May 3, 2026

coderabbitai Bot reviewed May 3, 2026

View reviewed changes

Comment thread CLAUDE.md

cristim changed the base branch from feat/multicloud-web-frontend to main June 9, 2026 15:52

cristim force-pushed the docs/issue-237-codify-pr-labeling branch from bbba22b to c3ae487 Compare June 19, 2026 15:43

cristim force-pushed the docs/issue-237-codify-pr-labeling branch from c3ae487 to 29c2c16 Compare July 10, 2026 14:30

cristim force-pushed the docs/issue-237-codify-pr-labeling branch from 29c2c16 to 194f63c Compare July 17, 2026 18:32

cristim force-pushed the docs/issue-237-codify-pr-labeling branch from 194f63c to 30a5a6f Compare July 19, 2026 19:00

cristim merged commit 95ecd19 into main Jul 19, 2026
19 checks passed

cristim deleted the docs/issue-237-codify-pr-labeling branch July 27, 2026 11:09

Conversation

cristim commented May 3, 2026 • edited by coderabbitai Bot Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Summary

Dogfooding

Test plan

Summary by CodeRabbit

Uh oh!

coderabbitai Bot commented May 3, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Walkthrough

Changes

Uh oh!

cristim commented May 3, 2026

Uh oh!

coderabbitai Bot commented May 3, 2026

Uh oh!

coderabbitai Bot left a comment

Choose a reason for hiding this comment

Uh oh!

Uh oh!

cristim commented May 3, 2026

Uh oh!

coderabbitai Bot commented May 3, 2026

Uh oh!

cristim commented May 19, 2026

CodeRabbit pass 1 — both findings addressed

Actionable (Major) — empty $LABELS guard

Nitpick (Quick win) — manual PR_NUM entry

Uh oh!

coderabbitai Bot commented May 19, 2026

Review summary

✅ Empty $LABELS guard (was Actionable/Major)

✅ PR_NUM auto-derivation (was Nitpick)

✅ ISSUE_NUM as intentional placeholder

✅ jq filter correctness

✅ Acceptance criteria from #237

Uh oh!

cristim commented May 20, 2026

Uh oh!

coderabbitai Bot commented May 20, 2026

Uh oh!

cristim commented May 20, 2026

Uh oh!

coderabbitai Bot commented May 20, 2026

Uh oh!

cristim commented May 21, 2026

Uh oh!

coderabbitai Bot commented May 21, 2026

Uh oh!

cristim commented May 21, 2026

Uh oh!

coderabbitai Bot commented May 21, 2026

Uh oh!

cristim commented May 22, 2026

Uh oh!

coderabbitai Bot commented May 22, 2026

Uh oh!

cristim commented May 22, 2026

CR finding verification - PR_NUM auto-derivation

Uh oh!

coderabbitai Bot commented May 22, 2026

Uh oh!

cristim commented May 22, 2026

Uh oh!

coderabbitai Bot commented May 22, 2026

Uh oh!

cristim commented Jun 4, 2026

Uh oh!

coderabbitai Bot commented Jun 4, 2026

Rate Limit Exceeded

Uh oh!

cristim commented Jun 6, 2026

Uh oh!

coderabbitai Bot commented Jun 6, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

cristim commented Jun 7, 2026

Uh oh!

coderabbitai Bot commented Jun 7, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

cristim commented May 3, 2026 •

edited by coderabbitai Bot

Loading

coderabbitai Bot commented May 3, 2026 •

edited

Loading

Actionable (Major) — empty `$LABELS` guard

Nitpick (Quick win) — manual `PR_NUM` entry

✅ Empty `$LABELS` guard (was Actionable/Major)

✅ `PR_NUM` auto-derivation (was Nitpick)

✅ `ISSUE_NUM` as intentional placeholder

✅ `jq` filter correctness

✅ Acceptance criteria from `#237`

coderabbitai Bot commented Jun 6, 2026 •

edited

Loading

coderabbitai Bot commented Jun 7, 2026 •

edited

Loading

coderabbitai Bot commented Jun 19, 2026 •

edited

Loading

coderabbitai Bot commented Jul 9, 2026 •

edited

Loading