新增實作規劃協調器技能#3
Conversation
![gemini-code-assist[bot]](https://avatars.githubusercontent.com/in/956858?s=60&v=4){kind=small}
There was a problem hiding this comment.
Code Review
This pull request introduces the implementation-orchestrator skill, a comprehensive framework for AI agents to manage software implementation workflows from initial intake to final delivery. The skill defines a structured process for context collection, implementation planning, complexity-based routing, and QA verification. The review feedback focuses on refining the technical artifacts for better AI reliability: specifically, replacing a mathematical formula with a list to avoid calculation errors, restructuring the complexity YAML to group levels with their rationales, and ensuring the QA checklist is explicitly derived from the implementation plan's criteria.
![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: 24dd77b09a
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".
| @@ -0,0 +1,561 @@ | |||
| --- | |||
| name: implementation-orchestrator | |||
| description: Turn engineering work into a reviewed execution baseline with a context pack, implementation plan, complexity routing, optional workstream breakdowns, execution briefs for coding agents or human executors, QA verification, and optional delivery updates. Use when the user wants to turn a bug, feature, refactor, review request, or technical task into an implementation plan, coordinate work across coding agents or human executors, or verify completed implementation against the approved plan. | |||
There was a problem hiding this comment.
Because skill selection is driven by this frontmatter description, including the broad phrase review request makes ordinary PR/commit review prompts match this orchestration workflow even though the workflow then produces context packs and implementation plans rather than review findings. In repos that also use a code-review agent, a user asking to review a commit or PR can be routed to the wrong skill; qualify this as review feedback that needs implementation planning, or remove the generic review-request trigger.
Useful? React with 👍 / 👎.
1 participant
將使用者提供的上下文資訊統籌為工程實作規劃，並將其轉化為：可供審查的計劃、複雜度評估、任務拆解、供編碼代理程式 (coding agents) 使用的執行摘要、品質保證 (QA) 驗證以及最終的交付產物。