workflow-practices
Workflow rules for task execution, respecting user code, and git commits. Use when planning or executing development tasks.
When & Why to Use This Skill
The workflow-practices skill establishes a robust framework for AI-driven development, ensuring task execution is structured, non-destructive, and aligned with professional engineering standards. By enforcing rules for codebase interaction and git management, it optimizes the collaboration between human developers and AI agents, maintaining code integrity and project consistency.
Use Cases
- Structured Feature Implementation: Automatically creates and follows a step-by-step todo list (Read -> Plan -> Execute) to ensure complex development tasks are handled logically and transparently.
- Safe Code Refactoring: Protects existing user-written logic and intentional code structures from being accidentally reverted or overwritten during AI-led refactoring sessions.
- Controlled Version Management: Prevents 'git noise' and accidental repository changes by strictly requiring explicit user permission before performing any git commits.
- Project Tooling Alignment: Ensures the AI agent respects and utilizes the specific project's existing environment, build tools, and configuration settings rather than introducing conflicting workflows.
| name | workflow-practices |
|---|---|
| description | Workflow rules for task execution, respecting user code, and git commits. Use when planning or executing development tasks. |
workflow-practices
Quick Start
- Read codebase -> Create todo list -> Check plan -> Execute
- NEVER revert user changes or overwrite intentional code
- Do NOT create git commits unless explicitly requested
- Respect project's tooling decisions
Key Principles
- Work WITH user changes, not against them
- Fully implement all functionality (no placeholders)
- Be concise, minimize prose
- If uncertain, say so
References
See rules.md for complete conventions.