Repository navigation
[SCR-195] Package the issue-standard templates - #196
Conversation
The rollout needs something to roll out. content-rabbit adopted the standard in its own repo and those files were reachable from nowhere else. Package them beside pr-standards-templates, which already works this way. The four forms and config.yml are the reviewed and merged files, copied rather than rewritten, so the fleet gets what was actually looked at. .agents/issues.md is the one that cannot be copied. It names the repo's analytics helper and its high-stakes paths, and neither can be derived from outside that repo. Inventing them would put a confident wrong answer into 88 places, so it ships as a stub carrying the parts that are identical everywhere and TODOs for the parts that are not. Its first line says it must be filled in before it is useful.
|
Warning Review limit reachedNext included review available in 29 minutes. View limit detailsLimit details: You’ve used the included review currently available. Your 106 included PR review attempts over the past 7 days set your current allowance at 1 review per hour. Your organization has reached its usage spending cap. Adjust your spending cap in the billing tab. Review configuration: ⚙️ Run configurationConfiguration used: Organization UI Review profile: CHILL Plan: Team Run ID: 📒 Files selected for processing (7)
Comment |
There was a problem hiding this comment.
Council review
Verified the one confirmed finding from council against the code (correctness, performance, security, and maintainability lenses returned nothing). No fixes pushed this round — see loop-guard note below.
Confirmed (Major): issue-standards-templates/bug.yml:44 still hardcodes the content-rabbit test workspace pooriaarabtestgrowth in the Environment field's placeholder. This is the same class of leak this PR has already fixed twice (repo links/analytics in 7998273, the d3k errors tool name in fb83c6a), but this instance survived both passes because the PR's own leak-check grep (grep -niE "content.rabbit|posthog|trackEvent|staging database" issues.md.stub) only scans issues.md.stub, not the rest of issue-standards-templates/. I confirmed it's still there and that no other file in the directory carries a similar leak.
I did not push a fix: this PR has already gone through 3 automated fix cycles and push access for this run is disabled, so this needs a human to (a) replace the placeholder with a generic value (e.g. Chrome 141, Growth plan, acme-marketing, staging) and (b) widen the verification grep to the whole issue-standards-templates/ directory so this class of leak stops slipping through one file at a time.
Other lenses checked and cleared:
- PR-standard compliance (
.github/pr-standards.jsonpresent): branchscr-195-package-the-issue-standard-templates, title[SCR-195] Package the issue-standard templates, body has exactly oneCloses #195,## What/## Why/## How I verified, ends inAssisted-by:. 453 counted lines / 7 files, well under the 500-line / 40-file caps. No violations. - Proof lens: the PR claims
issue-standards lint --dir issue-standards-templatespasses for all four forms using the checker from #192 (not yet on this branch's base, sincelintmode didn't exist there — the PR body is upfront about this). #192 has since merged tomainwith reallintsupport, so I checked its logic directly:lintTemplates()only compares each form'slabel:fields against required headings, and none of the three follow-up fix commits touched alabel:line (onlydescription:/prose text), so the PASS claim still holds today. I manually traced all four forms' field labels againstREQUIRED_HEADINGS/FORBIDDEN_HEADINGSinissue-standards.mjsand confirmed each form provides exactly the required headings (epic correctly omitsAcceptance criteria). The stub leak-check grep reproduces with no matches, confirming that specific claim — it just doesn't cover the file where the leak actually is (see above). - No test-coverage gap: this PR adds only template/markdown assets, no executable logic, so unit tests aren't applicable here.
🧑⚖️ Council
| Model | Lens | Result |
|---|---|---|
| GPT-5.6 (Codex) | correctness | no findings |
| Gemini 3 Pro | performance | no findings |
| Kimi K3 | security | no findings |
| Grok 4.5 | maintainability | no findings |
| GPT-5.6 (scope) | scope | confirmed — pooriaarabtestgrowth leak in bug.yml:44 |
| attributes: | ||
| label: Environment | ||
| description: Browser, plan, workspace, and whether this is local, staging or production. | ||
| placeholder: Chrome 141, Growth plan, pooriaarabtestgrowth, staging |
There was a problem hiding this comment.
Major — repo-specific leak survives two prior cleanup passes. This placeholder still names the content-rabbit test workspace pooriaarabtestgrowth. This PR's own "How I verified" leak-check grep only scans issues.md.stub, so it never touched this file. Please replace with a generic placeholder (e.g. acme-marketing) and widen the verification grep to cover all of issue-standards-templates/, not just the stub. Not pushing a fix here: this PR has already had 3 automated fix cycles and push is disabled for this run.
Closes #195
What
Packages the issue-standard templates beside
pr-standards-templates/: the four issue forms,config.yml, theAGENTS.mdmarker block, and a.agents/issues.mdstub.Why
The rollout in #183 needs something to roll out.
content-rabbitadopted the standard in pooriaarab/content-rabbit#1205, and those files were reachable from nowhere else.The forms and
config.ymlare the reviewed and merged files, copied rather than rewritten, so the fleet gets what was actually looked at rather than a second draft of it..agents/issues.mdis the one that cannot be copied. It names the repo's analytics helper and its high-stakes paths, and neither can be derived from outside that repo. Inventing them would put a confident wrong answer into 88 places, which is worse than an empty one, because an agent believes it. So it ships as a stub: the parts that are identical everywhere are verbatim, the parts that are not are TODOs, and the first line says it must be filled in before it is useful.How I verified
The forms still satisfy the checker that will judge issues filed through them:
Run with the checker from #192, because
lintis not onmainyet: onlyprecheckis. Onmaintoday the same command reports an unknown mode, which is correct and not a failure of these files.The stub names no repo but the one it is dropped into:
That grep found four leaks on the first draft, which is why it is written down here: the source file it was derived from is content-rabbit's, and the repo-specific parts do not announce themselves.
Size:
Not in this PR
issue-standards-rollout, the program that reads these and plans a repo. It is #183, split out because the two together are 614 counted lines and the cap is 500. The cap forced the split, which is what it is for.How it was built
The stub was drafted by a
museworker from content-rabbit's file. Review caught four repo-specific leaks it had carried over, including a line about test users and the staging database that is true only for content-rabbit, and a claim that sub-issues were "verified against the API" for a repo nobody had checked. Those are the exact failure this stub exists to prevent, so they are worth naming rather than quietly fixing.Assisted-by: claude-personal-1:claude-opus-5