Skip to content

feat: add Q1 submission gates for statistics, peer review, and declarations - #8

Merged
mxuanvan02 merged 1 commit into
mainfrom
feat/q1-submission-gates
Sep 6, 2026
Merged

mxuanvan02 merged 1 commit into
mainfrom
feat/q1-submission-gates

Conversation

@mxuanvan02

Copy link
Copy Markdown
Owner

What

Three genre contracts this repository had no coverage for. A probe across all
24 existing references found zero files mentioning reviewer responses, cover
letters, confidence intervals, effect size, p-values, multiplicity, statistical
power, baseline parity, or generative-AI disclosure — each a routine desk-reject
cause at a leading journal.

File Owns
references/quantitative-reporting-standard.md denominators, intervals, effect size, multiplicity, baseline parity, small-sample t vs normal
references/revision-response-genres.md the response letter as an artifact separate from the manuscript; verdict discipline; no capitulation
references/submission-integrity-declarations.md authorship, ethics, funding, availability, AI disclosure as author-owned facts that are never inferred

Registration

22 new blocking codes registered in both quality-rubric.md and
writing-failure-taxonomy.md, plus three new rubric gate sections. 11 further
codes reuse existing entries rather than duplicating them.

Proof the gate can go red

Per references/skill-repository-maintenance.md §3, the new contract test was
mutated five ways and each mutation was verified present before the run and
absent after restore:

  1. rubric gate heading deleted → caught
  2. blocking code removed from rubric → caught
  3. code removed from taxonomy only → caught
  4. reference unlinked from SKILL.md (both occurrences) → caught
  5. duplicate section ownership reintroduced → caught

The first attempt at mutation 4 reported a false MISSED because SKILL.md
references each file twice and only one occurrence was replaced — the harness
now asserts the construct is fully absent before trusting a negative result.

Verification

All five CI steps green locally: validator rc=0, 25 repository tests, and
29 + 13 + 30 scanner tests (97 total).

Version 3.5.2 → 3.6.0.

…ations

Three genre contracts the repository had no coverage for, each a common
desk-reject cause at a leading journal:

- quantitative-reporting-standard.md: denominators, intervals, effect size,
  multiplicity, baseline parity, and the small-sample t/normal decision.
- revision-response-genres.md: the response letter as an artifact separate
  from the manuscript, with verdict discipline and no capitulation.
- submission-integrity-declarations.md: authorship, ethics, funding,
  availability, and generative-AI disclosure as author-owned facts that
  must never be inferred.

Registers 22 new blocking codes in quality-rubric.md and
writing-failure-taxonomy.md, adds three rubric gate sections, and pins the
whole set with a contract test proven red by five mutations (deleted gate
heading, removed rubric code, taxonomy-only removal, reference unlinked from
SKILL.md, duplicate section ownership reintroduced).

Version 3.6.0.
@chatgpt-codex-connector

Copy link
Copy Markdown

You have reached your Codex usage limits for code reviews. You can see your limits in the Codex usage dashboard.

@mxuanvan02
mxuanvan02 merged commit be0c16b into main Sep 6, 2026
2 checks passed
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant