A/Bテストや実験の設計・実装を支援するスキル。 「A/Bテストを設計して」「スプリットテストしたい」「仮説を立ててテストしたい」「バリアントを比較」等のリクエストで発動。 トラッキング実装は analytics-tracking を参照。
aiagent-lesson-runner
ai-agent-campのレッスンをCodexで開始・進行するスキル。 「レッスン開始」「次のレッスン」「start-0-1を始めたい」「Codexでレッスン」「スラッシュコマンドのレッスン」等のリクエストで発動。
インストール方法を見る含まれるファイル(4)
- SKILL.md3.2 KB
- agents/openai.yaml254 B
- SKILL.en.md2.9 KB
- SKILL.es.md3.3 KB
SKILL.md(原文)
インストールする前に、エージェントに与えられる指示の中身を確認できます。
トリガーワード
「レッスン開始」「次のレッスン」「レッスン実行」「スラッシュコマンド」「Codexレッスン」
AI Agent Lesson Runner
Use this skill to reproduce the start-* workflow in Codex.
Inputs
- A lesson id such as
start-0-1
Workflow
- Before running any lesson on a freshly cloned repo, run
python3 tools/scripts/verify_integrity.py. If it reports the origin is not official (github.com/minicoohei/ai-agent-camp) or required files are missing, stop and ask the user to review the diff against upstream/main. - Validate that the lesson id matches
^start-\d+-\d+$. Reject any other value. - Check that
.cursor/commands/lesson/<lesson-id>.mdexists before reading it. - Open
.cursor/commands/lesson/<lesson-id>.md. - Extract the lesson goal, prerequisites, checkpoints, command references, and referenced files.
- Resolve the curriculum source in this order:
- frontmatter
chapterwhen it points to acourses/aiagent/**/chapter*.yaml - sibling
practice/andfinal/docs next to that chapter - if
chapteris missing, infer the best source fromcourses/lessons.manifest.yaml, the lesson URL, and thestart-X-Ymodule number
- frontmatter
- Treat embedded
AskQuestion/AskUserQuestionJSON as a conversation blueprint. In Codex, convert it into concise numbered or bulleted options in normal chat instead of pretending the Cursor UI exists. - For
/setup-start,/setup-github,/check-setup, and similar setup flows, do not tell the user to run the Cursor slash command literally. Execute or describe the underlying checks and split GUI-required steps from AI-executable steps. - If the lesson touches Git, secrets, MCP, or external APIs, tell the user which safety doc to read first and verify prerequisites before attempting the task.
- Guide the user through:
- prerequisite check
- files to read
- actions to perform
- completion criteria
- next recommended lesson
Required References
.cursor/commands/lesson/start-*.md- matching
courses/aiagent/**/chapter*.yamlwhen available - sibling
practice/orfinal/docs for the resolved chapter courses/lessons.manifest.yamlas the fallback lookup table when a lesson markdown is missingchapter
Safety
- If the lesson implies Git or environment changes, also consult
docs/codex-safety.md. - Never pretend the user can run the Cursor markdown command file directly in Codex.
- Never ask the user to paste secret values into chat. Reuse the repo credential workflow instead.
Expected Output
- Lesson summary
- Ordered next actions
- Relevant files
- Done criteria
- Suggested follow-up lesson
- When the lesson contains structured choices, present the Codex-friendly options inline in the response
レビュー
まだレビューはありません。使ってみた感想をお寄せください。
同じリポジトリのスキル
概要と使いどころ
マルチエージェントシステムのアーキテクチャ設計ツールキット。 「エージェントを設計して」「マルチエージェント構成」「エージェントのアーキテクチャ」「オーケストレーション設計」等のリクエストで発動。
ai-agent-campのローカル環境セットアップを確認するスキル。 「セットアップ確認」「環境チェック」「初期設定できてる?」「インストール確認」「依存関係チェック」等のリクエストで発動。
ai-agent-campのスラッシュコマンドをCodexでルーティングするスキル。 「/start-0-1を実行」「スラッシュコマンドを使いたい」「コマンドルーティング」「Cursorのコマンドを使いたい」等のリクエストで発動。
ai-agent-campの環境変数・認証情報を安全に管理するスキル。 「APIキーを設定」「.envを管理」「環境変数の設定」「credential managerを使いたい」「秘密情報の管理」等のリクエストで発動。
ai-agent-campリポジトリの案内・オリエンテーションスキル。 「リポジトリ案内」「次のレッスンは?」「どこから始める」「ツールの違い」「aiagent概要」等のリクエストで発動。