Repository navigation
docs: Chat / Work用とCodex用のカスタム指示を分離 - #173
Conversation
Codex Review SummaryThis comment shows the latest Codex review activity on this pull request.
ℹ️ About Codex in GitHubYour team has set up Codex to review pull requests in this repo. Reviews are triggered when you
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[bot]](https://avatars.githubusercontent.com/in/1144995?s=60&v=4){kind=small}
There was a problem hiding this comment.
💡 Codex Review
Here are some automated review suggestions for this pull request.
Reviewed commit: d1b719f864
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".
| * 調査結果や技術的な説明についても、単に情報を列挙するのではなく、事実関係や理由、注意点のつながりが読み手に自然に伝わるように説明してください。 | ||
| ``` | ||
|
|
||
| ## Codex用のカスタム指示 |
There was a problem hiding this comment.
This new level-2 heading remains active until the next ## heading at line 588, so the pre-existing sections about AdGuard with the ChatGPT Android app, general prompt workflow, and role prompting are now structurally grouped under “Codex用のカスタム指示.” In rendered outlines and section-based navigation, this incorrectly presents general ChatGPT guidance as Codex-specific content; add an equivalent-level boundary after the Codex code block or give those existing blocks their own heading.
AGENTS.md reference: AGENTS.md:L6-L8
Useful? React with 👍 / 👎.
| 【基本方針】 | ||
|
|
||
| - 正確性を最優先してください。 | ||
| - 作業開始前に、README、AGENTS.md、CONTRIBUTING、関連ドキュメント、対象ファイル、既存テスト、GitHub Actions等を確認し、リポジトリの構造・目的・既存ルールを把握してください。 |
There was a problem hiding this comment.
This global instruction requires Codex to inspect README, AGENTS.md, CONTRIBUTING, related documentation, existing tests, and GitHub Actions before every task, including a one-line typo fix. OpenAI's current Codex guidance specifically identifies mandatory stacks of documentation or full-repository mapping before every edit as excessive because it consumes context and slows work; make these checks conditional on which parts of the repository the task affects.
Useful? React with 👍 / 👎.
| - 問題がある場合は原因を修正 | ||
| - 検証を無効化して通過させない | ||
| - 問題が解消してからコミット・PR・マージする |
There was a problem hiding this comment.
When the repository already has an unrelated failing test, lint check, or Action, these unconditional steps direct Codex to fix that problem and withhold the requested commit until it is resolved. That either expands a focused task into unrelated files or blocks delivery for a pre-existing/environmental failure; limit the repair requirement to failures caused by the requested change and require unrelated failures to be reported instead.
AGENTS.md reference: AGENTS.md:L6-L8
Useful? React with 👍 / 👎.
1 participant
「ChromeOS & Android Optimization Guide.md」のChatGPT プロンプト内で、既存のカスタム指示をChat / Work用として明示し、提供された指示全文を独立したCodex用セクションに追加します。既存の指示本文・注釈・参考リンクと対象外の内容は保持しています。
検証：Markdown lint（21ファイル・指摘0件）、関連テスト13件、既存の内容保全ガード、git diff --checkが成功。追加部分と見出し変更を取り除くと元ファイルに完全一致することも確認済みです。Versionは既存Workflowによるmain反映後の自動同期に従います。