Cursor rule

1---
2description: Comprehensive PR review using specialized agents
3alwaysApply: true
4---
5 
6# Comprehensive PR Review
7 
8Run a comprehensive pull request review using multiple specialized agents, each focusing on a different aspect of code quality.
9 
10**Review Aspects (optional):** "$ARGUMENTS"
11 
12## Review Workflow:
13 
141. **Determine Review Scope**
15   - Check git status to identify changed files
16   - Parse arguments to see if user requested specific review aspects
17   - Default: Run all applicable reviews
18 
192. **Available Review Aspects:**
20 
21   - **comments** - Analyze code comment accuracy and maintainability
22   - **tests** - Review test coverage quality and completeness
23   - **errors** - Check error handling for silent failures
24   - **types** - Analyze type design and invariants (if new types added)
25   - **code** - General code review for project guidelines
26   - **simplify** - Simplify code for clarity and maintainability
27   - **all** - Run all applicable reviews (default)
28 
293. **Identify Changed Files**
30   - Run `git diff --name-only` to see modified files
31   - Check if PR already exists: `gh pr view`
32   - Identify file types and what reviews apply
33 
344. **Determine Applicable Reviews**
35 
36   Based on changes:
37   - **Always applicable**: code-reviewer (general quality)
38   - **If test files changed**: pr-test-analyzer
39   - **If comments/docs added**: comment-analyzer
40   - **If error handling changed**: silent-failure-hunter
41   - **If types added/modified**: type-design-analyzer
42   - **After passing review**: code-simplifier (polish and refine)
43 
445. **Launch Review Agents**
45 
46   **Sequential approach** (one at a time):
47   - Easier to understand and act on
48   - Each report is complete before next
49   - Good for interactive review
50 
51   **Parallel approach** (user can request):
52   - Launch all agents simultaneously
53   - Faster for comprehensive review
54   - Results come back together
55 
566. **Aggregate Results**
57 
58   After agents complete, summarize:
59   - **Critical Issues** (must fix before merge)
60   - **Important Issues** (should fix)
61   - **Suggestions** (nice to have)
62   - **Positive Observations** (what's good)
63 
647. **Provide Action Plan**
65 
66   Organize findings:
67```markdown
68   # PR Review Summary
69 
70   ## Critical Issues (X found)
71   - [agent-name]: Issue description [file:line]
72 
73   ## Important Issues (X found)
74   - [agent-name]: Issue description [file:line]
75 
76   ## Suggestions (X found)
77   - [agent-name]: Suggestion [file:line]
78 
79   ## Strengths
80   - What's well-done in this PR
81 
82   ## Recommended Action
83   1. Fix critical issues first
84   2. Address important issues
85   3. Consider suggestions
86   4. Re-run review after fixes
87```
88 
89## Usage Examples:
90 
91**Full review (default):**
92```
93/pr-review-toolkit:review-pr
94```
95 
96**Specific aspects:**
97```
98/pr-review-toolkit:review-pr tests errors
99# Reviews only test coverage and error handling
100 
101/pr-review-toolkit:review-pr comments
102# Reviews only code comments
103 
104/pr-review-toolkit:review-pr simplify
105# Simplifies code after passing review
106```
107 
108**Parallel review:**
109```
110/pr-review-toolkit:review-pr all parallel
111# Launches all agents in parallel
112```
113 
114## Agent Descriptions:
115 
116**comment-analyzer**:
117- Verifies comment accuracy vs code
118- Identifies comment rot
119- Checks documentation completeness
120 
121**pr-test-analyzer**:
122- Reviews behavioral test coverage
123- Identifies critical gaps
124- Evaluates test quality
125 
126**silent-failure-hunter**:
127- Finds silent failures
128- Reviews catch blocks
129- Checks error logging
130 
131**type-design-analyzer**:
132- Analyzes type encapsulation
133- Reviews invariant expression
134- Rates type design quality
135 
136**code-reviewer**:
137- Checks CLAUDE.md compliance
138- Detects bugs and issues
139- Reviews general code quality
140 
141**code-simplifier**:
142- Simplifies complex code
143- Improves clarity and readability
144- Applies project standards
145- Preserves functionality
146 
147## Tips:
148 
149- **Run early**: Before creating PR, not after
150- **Focus on changes**: Agents analyze git diff by default
151- **Address critical first**: Fix high-priority issues before lower priority
152- **Re-run after fixes**: Verify issues are resolved
153- **Use specific reviews**: Target specific aspects when you know the concern
154 
155## Workflow Integration:
156 
157**Before committing:**
158```
1591. Write code
1602. Run: /pr-review-toolkit:review-pr code errors
1613. Fix any critical issues
1624. Commit
163```
164 
165**Before creating PR:**
166```
1671. Stage all changes
1682. Run: /pr-review-toolkit:review-pr all
1693. Address all critical and important issues
1704. Run specific reviews again to verify
1715. Create PR
172```
173 
174**After PR feedback:**
175```
1761. Make requested changes
1772. Run targeted reviews based on feedback
1783. Verify issues are resolved
1794. Push updates
180```
181 
182## Notes:
183 
184- Agents run autonomously and return detailed reports
185- Each agent focuses on its specialty for deep analysis
186- Results are actionable with specific file:line references
187- Agents use appropriate models for their complexity
188- All agents available in `/agents` list
189