[kyle-sexton]

No related issue: owner-approved design (org issue and PR template interview, Q34 and Q36-Q38, 2026-11-11)

Summary

The org issue forms are now generated in melodic-software/standards from one issue-body contract, and the standards sync delivers that contract here at docs/contracts/issue-body/contract.json (#15429). This PR makes the skills that write or check issue bodies read their skeletons from it, so a skill-filed issue and a form-filed issue carry the same headings.

Fix

• Fan-out: scripts/sync-shared-copies.sh copies a JSON canonical byte for byte (JSON has no comment syntax for a generated-file header) and generates plugins/bugs/reference/issue-body-contract.json and plugins/work-items/reference/issue-body-contract.json; shared-copies.txt and cross-plugin-source-registry.txt register the cluster. Tests cover the JSON copy and drift.
• /work-items:track add: the body's sections come from the contract type matching the resolved issue type (bug, feature, task), or agent-ready; the inline Context / Proposed work / Acceptance criteria / References skeleton is gone. Metadata and Recurring stay as appended sections.
• Agent brief (reference/agent-brief.md): fields are the contract's agent-ready type, written as ### headings under ## Agent Brief, the same level the agent-ready form renders.
• /work-items:decompose: its required sections are the contract's decomposed-item type; the contract wins where they disagree.
• /work-items:triage: step 1 checks an issue body against the contract with the new scripts/check-body.mjs (accepts ## and ###, treats a form's _No response_ as empty, ignores fenced code and HTML comments); steps in context/body-check.md.
• /bugs:write: the template's sections are the contract's bug fields in order (Severity and Suggested fix location become sections, matching the form).
• /code-tidying:tidy: the deferred-item body follows the task type's required headings.
• reference/dogfood-filing.md points at the contract instead of restating the skeleton.

Verification

• node --test plugins/work-items/skills/triage/scripts/check-body.test.mjs: 8 pass; the CLI on a form-shaped bug body against the real contract returns no gaps for bug and names the three missing task fields.
• bash scripts/sync-shared-copies.test.sh: 99 pass; sync-shared-copies.sh --check: all 804 copies match.
• scripts/check-cross-plugin-source-drift.sh --check: no unregistered or drifted clusters.
• scripts/check-changed-skills.sh: 5 skills pass (triage held under its 500-line cap by moving the steps to a context file).
• Pre-push scripts/check-affected.sh: passes.

Related

• feat(issue-body-contract): add the issue-body contract and generate the org issue forms from it standards#986 and fix(issue-body-contract): deliver the plugin contract to one root path standards#988 (the contract, generator and delivery)
• chore: sync standards components .github#237 and ci: check the issue forms against the issue-body contract .github#238 (the synced forms and their drift check)
• chore: sync standards components #15429 (the sync that delivered the canonical)

🤖 Generated with Claude Code

https://claude.ai/code/session_01VQb9Q2ceZLXhTR4d1vkTCe

[chatgpt-codex-connector]

Codex Review Summary: ✅ Completed

This comment shows the latest Codex review activity on this pull request.

Review	Status	Commit	Review trigger
📝 Code Review	✅ Completed 2026-11-11T22:50:37.550031Z	0c789af	Draft marked ready

ℹ️ About Codex in GitHub

Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you

• Open a pull request for review
• Mark a draft as ready
• Comment "@codex review" or "@codex security review".

Codex reacts with 👀 while any review is running, comments if it has suggestions, and reacts with 👍 once all reviews finish with no findings.

[chatgpt-codex-connector]

💡 Codex Review

Here are some automated review suggestions for this pull request.

Reviewed commit: 0c789afd8b

ℹ️ About Codex in GitHub

Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you

• Open a pull request for review
• Mark a draft as ready
• Comment "@codex review".

If Codex has suggestions, it will comment; otherwise it will react with 👍.

Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".

[github-actions]

claude-security-review reviewed 1afd90696ff567d2d313e01893532900afa8ab64: read 4 of 4 changed file(s); 0 finding(s), 0 as inline comments.

Security review: PR #15453 (incremental, cb3a459 → 1afd906)

I found no security issues.

critical-findings: 0

Scope: I read every hunk of the 4 files in the incremental scope. None of them matches a TRUSTED PATHS glob. The trusted plugins/work-items/scripts/charge-words.json is not touched.

Blocking only: this lane already reviewed an earlier head, so this review would have reported only CRITICAL or IMPORTANT findings. There were none.

What I checked

• plugins/work-items/reference/agent-brief.md and plugins/work-items/skills/track/actions/add.md: "contract order" / "the contract's" now read "issue-body contract order" / "the issue-body contract's". This only names the contract. It removes, narrows or softens no instruction, so the instruction-surface lens has nothing to judge.
• scripts/vocabulary-senses.json: adds one exemption phrase, \bissue-body contract\b, to the vocabulary checker's contract sense. That checker is a wording lint. It is not a trust control, a merge hold or a gate on credentials. The regex is a fixed literal with word boundaries, so it brings no ReDoS risk.
• scripts/test_check_vocabulary.py: adds one test row for the new exemption. It is test-only code and takes no outside input.
• Trust controls, tokens and manifests: the change touches no workflow, token request, kill switch, trusted-actor list or plugin manifest.
• Spend: none.

[github-actions]

claude-review: 1 inline finding(s) on 0c789afd8ba8227ce5b976d398de62e2e111e52e; the summary is in this lane's pull request comment.

claude-lane: claude-review

[github-actions]

claude-review reviewed 1afd90696ff567d2d313e01893532900afa8ab64: read 4 of 4 changed file(s); 0 finding(s), 0 as inline comments.

Code review (incremental, blocking only): cb3a4598..1afd9069

This lane already reviewed earlier heads of this PR, so this pass reports blocking findings only. I read every hunk of the 4 files in .claude-lane/incremental-files.txt. No blocking findings.

What the push changes. It adds a vocabulary-check exemption for the phrase "issue-body contract" and rewrites two prose lines to use that name:

• scripts/vocabulary-senses.json adds the exemption to the contract-brief sense.
• scripts/test_check_vocabulary.py adds a matching phrase(...) test row. Its regex is the same as the JSON entry's, and its sample line matches it.
• plugins/work-items/reference/agent-brief.md:34 and plugins/work-items/skills/track/actions/add.md:55 now say "issue-body contract" where they said "contract".

Does the exemption clear the lines it targets? I read scripts/check-vocabulary.py:190-197. A phrase exemption clears the whole line it appears on, not only the matched words. So:

• add.md:55 keeps several bare uses of "contract": "in contract order", "the contract type" and "the contract does not hold". It also matches the near rule's \bbrief\b through agent-brief.md. The one new "issue-body contract" on the line clears it all.
• The other lines of agent-brief.md that mention "contract" either already contain the exempt phrase (line 31) or have no "brief" on them (line 32, issue-body-contract.json).

Prior findings and settled threads. The ### sub-heading thread on check-body.mjs is settled (lines changed since resolution: no), and this push doesn't touch that file. There is nothing to post again.

Closing issue: the PR context names none, so there are no acceptance criteria to check.

Manifests: this push changes no plugin.json or marketplace.json.

Verification: this run had no shell, so I did not run scripts/test_check_vocabulary.py or the vocabulary gate. The checks above come from reading the code only.