Structured Workflow Engine MCP Server

# Code Review Prompt (v2) ## 🎯 Goal Audit the finished code for quality, security, performance, and standards, then hand back a crisp report—no fluff. ## 📥 Context (ask if missing) 1. **Code Branch / PR** – where’s the diff? 2. **Requirements / Specs** – doc or ticket link. 3. **Known Constraints** – perf targets, security mandates, style guide, etc. 4. **Deadline** – when do devs need feedback? ## 🚦 Skip if - Only trivial config tweaks **or** review already done. ## 🔍 Checklist - **Functional** - [ ] Implements all requirements & edge cases - **Quality** - [ ] Readable, DRY, follows style guide - [ ] No dead code / duplication - **Security** - [ ] Input validation / sanitization - [ ] Secrets handled via env/secret manager - [ ] AuthN/Z correct - **Performance** - [ ] No obvious bottlenecks or N+1 queries - [ ] Memory & CPU use sane - **Error Handling** - [ ] Graceful exceptions, helpful logs - **Testing** - [ ] Adequate unit / integration coverage - [ ] Tests pass and are maintainable ## 📤 Output **File:** `docs/planning/[feature-name]-code-review.md` Structure: 1. **Summary** – 🚦 Approved / Approved-with-changes / Needs-rework 2. **Issue Table** | Severity | File/Line | Issue | Recommendation | |----------|-----------|-------|----------------| | Critical | `auth.go:42` | SQL injection risk | Use prepared stmt | 3. **Security Findings** – bullets 4. **Performance Notes** – bullets 5. **Quality & Style** – highlights + nitpicks 6. **Test Coverage** – % plus missing cases 7. **Next Steps** – who fixes what by when ## ➡️ Response Flow ```mermaid flowchart LR U[User] -->|PR ready| A[Review Engine] A --> B{Need more context?} B -- Yes --> C[Ask for branch / specs] B -- No --> D[Run review] D --> E[Write code_review.md]