chore: strengthen AI disclosure review policy - #3065
Conversation
|
No actionable comments were generated in the recent review. 🎉 ℹ️ Recent review info⚙️ Run configurationConfiguration used: Path: .coderabbit.yaml Review profile: CHILL Plan: Pro Plus Run ID: 📒 Files selected for processing (4)
📝 WalkthroughWhat this PR solvesThis PR standardizes disclosure of substantive generative-tool use. It requires contributors to identify the tool and scope, and to add matching It also adds advisory CodeRabbit checks for missing or inconsistent disclosures and reports concrete review-relevant risks. The merger evaluates the final review path, and a maintainer makes the final determination. Source of truthThis PR extends the existing contribution policy and pull request template. It does not create a parallel application or review path. The CodeRabbit configuration enforces the same disclosure rules in an advisory form. Solution scope and complexityThe changes form the smallest coherent solution for:
The added configuration is necessary to detect inconsistent disclosures before merge. Automatic reviewer assignment remains disabled. Simplification opportunitiesNo deletion or simplification is apparent without weakening disclosure enforcement or review guidance. The configuration already removes redundant Risks and validationThe changes affect contributor workflow and repository governance. They require explicit AI-use declarations, The PR reports CodeRabbit schema validation, formatting checks, whitespace checks, and confirmation that contributor policies do not mention CodeRabbit or automated review bots. The final status of these checks is not independently verified here. Review-relevant risks
The person performing the merge reviews the final diff. A maintainer makes the final determination. WalkthroughThe PR adds AI-use disclosure requirements to templates and contribution guides. It adds a warning check for disclosure and commit-trailer consistency, disables automatic reviewer assignment, and updates fast-path approval rules in English and Chinese documentation. ChangesAI disclosure governance
Estimated code review effort: 3 (Moderate) | ~20 minutes Merge Risk: ⚪ Minimal · up to This PR updates contribution guidance, the pull request template, and review configuration for AI-use disclosure; no actionable merge-blocking risk remains beyond normal checks and review. Sequence Diagram(s)sequenceDiagram
participant PRDescription
participant CommitMessages
participant AIUseDisclosureCheck
PRDescription->>AIUseDisclosureCheck: disclosure declaration and tool scope
CommitMessages->>AIUseDisclosureCheck: Generated-by trailers
AIUseDisclosureCheck-->>PRDescription: validation result and corrective guidance
Possibly related PRs
Suggested reviewers: 🚥 Pre-merge checks | ✅ 6✅ Passed checks (6 passed)
✨ Finishing Touches🧪 Generate unit tests (beta)
Comment |
7f161fd to
8c877d9
Compare
Require explicit PR-level AI-use declarations while preserving the distinction between substantive assistance and material AI-authored content. Add advisory review triage for attribution and fast-path eligibility, with final authority remaining with maintainers. Generated-by: Codex
8c877d9 to
e208db8
Compare
CodeRabbit plan requirementThis PR configures the custom pre-merge check
Because Maka is a public Apache-2.0 open-source project, CodeRabbit's current policy indicates that this repository receives the Pro+ feature set through the OSS tier, so the custom check should be available without a paid Pro+ subscription. OSS-specific review rate limits still apply, and this conclusion should be revisited if CodeRabbit changes its OSS policy. |
Astro-Han
left a comment
There was a problem hiding this comment.
I think the AI disclosure part is in good shape. The fast-path part can be simpler.
Right now the same decision appears in the PR template, the CodeRabbit summary, and the final maintainer action. Only the last one can be based on the final diff and checks. I suggest:
- Keep the policy in
CONTRIBUTING.md, withCONTRIBUTING.zh-CN.mdas its translation. - Keep the AI-use declaration in the PR template, but remove the
Review pathsection andFast-path rationale. - Have CodeRabbit warn when a change appears to affect user-visible behavior, public contracts, security, licensing, releases, or governance. It should not classify a PR as a fast-path candidate.
- For a fast-path merge, the person performing the merge should leave one comment after reviewing the final diff, explaining why the change is low impact, reversible, and outside the protected areas. If another commit is pushed afterward, the comment needs to be posted again. GitHub already records CI status, so the comment does not need to repeat it.
- Keep the AI disclosure check focused on the selected declaration, named tools and scope, matching
Generated-bytrailers, and avoiding guesses based on writing or code style. Other policy reminders can stay inCONTRIBUTING.md. - Remove
suggested_reviewers: true, since it is already the default. Keepauto_assign_reviewers: falseas an explicit boundary so the bot cannot assign human reviewers.
This leaves a simpler division of responsibility: the author reports facts, bots point out risks, and a human makes the review decision.
AI-assisted review: Codex checked the PR diff and current CodeRabbit schema and documentation. Claude and K3-256k provided independent read-only consultations. The human contributor approved posting this comment. This does not count as independent human review.
简体中文
我觉得 AI 披露部分已经比较完整，fast path 部分还可以再简单一些。
目前同一个决定分别出现在 PR 模板、CodeRabbit 摘要和维护者最终操作中。只有最后一个环节能基于最终 diff 和检查结果作出判断。建议：
- 政策正文只放在
CONTRIBUTING.md，CONTRIBUTING.zh-CN.md作为对应翻译。 - PR 模板保留 AI 使用声明，删除
Review path和Fast-path rationale。 - CodeRabbit 只在改动可能涉及用户可见行为、公开契约、安全、许可、发布或治理时提醒需要关注，不判断 PR 是否属于 fast-path candidate。
- 采用 fast path 时，由实际执行合并的人在审阅最终 diff 后发一条评论，说明改动影响较低、容易回退且不涉及受保护领域。如果评论后又有新的 commit，需要重新评论。CI 状态由 GitHub 记录，评论中无需重复。
- AI 披露检查只核对声明选项、工具及参与范围、匹配的
Generated-bytrailer，并禁止根据写作或代码风格猜测是否使用 AI。其他政策提醒留在CONTRIBUTING.md。 - 删除已经是默认值的
suggested_reviewers: true，保留auto_assign_reviewers: false，明确机器人不能自动指派人工 reviewer。
这样职责更清楚：作者陈述事实，机器人提示风险，由人决定审查路径。
本次 Review 使用了 AI 辅助：Codex 核对了 PR diff、当前 CodeRabbit schema 和官方文档；Claude 与 K3-256k 分别进行了只读咨询。人类贡献者确认发送本评论。这不属于独立人工 Review。
Keep authors focused on factual AI-use disclosure while reserving fast-path decisions for the final revision. CodeRabbit now flags review-relevant risks without classifying or authorizing the review path. Generated-by: Codex
|
@Astro-Han Thanks — addressed in commit 8a0a072.
This leaves authors responsible for reporting facts, CodeRabbit responsible for surfacing risks and reviewer suggestions, and maintainers responsible for the final review path. 简体中文感谢，已在 commit 8a0a072 中处理：
现在的职责划分是：作者报告事实，CodeRabbit 提示风险并推荐 reviewer，维护者决定最终审查路径。 |
Astro-Han
left a comment
There was a problem hiding this comment.
Reviewed the latest revision and agree with the resulting policy and division of responsibilities. The previous concerns have been addressed: authors disclose AI use, CodeRabbit surfaces evidence-based risks without choosing the review path, and the merger evaluates the final revision.\n\nAI-assisted review: Codex compared the latest diff with the prior review findings. I reviewed the result and own this approval.
English
Summary
Make AI-use disclosure explicit and reviewable without coupling the contributor policy to any review bot.
Add structured PR declarations, advisory CodeRabbit checks for missing or inconsistent disclosure, review-relevant risk flags, and native reviewer suggestions. CodeRabbit does not classify a PR as a fast-path candidate; the person performing the merge evaluates the final revision, and a maintainer makes the final determination.
Refs #3051
Verification
.coderabbit.yamlagainst the current CodeRabbitschema.v2.jsonnpm run format:checkgit diff --checkCONTRIBUTING.mdand its Chinese translation do not mention CodeRabbit or an automated review botAI use
Select exactly one:
Tool(s) and scope: Codex researched CodeRabbit's documented capabilities and drafted the policy, PR template, review configuration, and review-follow-up revisions. The human contributor reviewed the scope and directed the policy boundary.
Checklist
Does this PR entail a change in behavior?
简体中文
摘要
让 AI 使用披露变得明确且可审查，同时避免贡献政策依赖或提及任何 review bot。
增加结构化 PR 声明、用于发现披露缺失或矛盾的 CodeRabbit 提醒、与审查相关的风险提示，以及原生 reviewer 推荐。CodeRabbit 不再将 PR 分类为 fast-path candidate；由实际执行合并的人评估最终版本，并由维护者作出最终判断。
关联 #3051
验证
schema.v2.json校验了.coderabbit.yamlnpm run format:checkgit diff --checkCONTRIBUTING.md及其中文版本均未提及 CodeRabbit 或自动 review botAI 使用
仅选择一项：
工具及参与范围：Codex 调研了 CodeRabbit 的公开配置能力，并起草了政策、PR 模板、review 配置及后续 review 修订。人类贡献者审阅了范围并确定了政策边界。
检查清单
本 PR 是否涉及行为变更？