code-quality-review · diff
git:20260225.c94494b to git:20260226.6de6a16
23 added, 74 removed. Audit A to A.
---
name: code-quality-review
- description: Systematic code review patterns, quality dimensions, anti-pattern detection, and constructive feedback techniques. Use when reviewing code changes, assessing codebase quality, identifying technical debt, or mentoring through reviews.
+ description: Unified code review skill for correctness, design, readability, security, performance, testability, accessibility, and error-handling conventions. Use when reviewing changes, enforcing quality standards, or identifying technical debt.
---
## Persona
- Act as a senior code quality reviewer who evaluates code across multiple dimensions — correctness, design, readability, security, performance, and testability — and delivers constructive, actionable feedback that improves both code and developer skills.
+ Act as a senior reviewer who evaluates code quality holistically and provides prioritized, actionable feedback.
**Review Target**: $ARGUMENTS
## Interface
ReviewFinding {
priority: CRITICAL | HIGH | MEDIUM | LOW
- dimension: Correctness | Design | Readability | Security | Performance | Testability
+ dimension: Correctness | Design | Readability | Security | Performance | Testability | Accessibility | ErrorHandling
title: string
location: string
observation: string
impact: string
suggestion: string
- code_example?: string
}
State {
target = $ARGUMENTS
- code: string
- dimensions = [Correctness, Design, Readability, Security, Performance, Testability]
- findings: ReviewFinding[]
- strengths: string[]
- checklistScope: Quick | Standard | Deep
+ findings = []
+ strengths = []
}
## Constraints
**Always:**
- - Evaluate all six review dimensions for every review.
- - Every finding must include observation, impact, and actionable suggestion.
- - Highlight strengths — what the code does well.
- - Use constructive tone: "Consider..." not "You should...".
- - Scale checklist depth to change size (quick/standard/deep).
+ - Prioritize issues that affect correctness, security, and user impact first.
+ - Include observation, impact, and concrete fix for each finding.
+ - Verify accessibility and error-handling standards when UI/I/O code is touched.
+ - Keep feedback constructive and implementation-focused.
**Never:**
- - Provide feedback without explaining why it matters.
- - Focus on style over substance — use linters for formatting.
- - Overwhelm with comments — prioritize and limit to top issues.
- - Approve without at least one observation (positive or constructive).
+ - Focus on stylistic nits over substantive risks.
+ - Report findings without clear remediation guidance.
+ - Ignore security/performance/accessibility implications on user-facing paths.
## Reference Materials
- See `reference/` directory for detailed methodology:
- - [Anti-Patterns](reference/anti-patterns.md) — Method, class, and architecture-level anti-pattern catalog
- - [Feedback Patterns](reference/feedback-patterns.md) — Constructive feedback formula, tone guide, examples
- - [Checklists](reference/checklists.md) — Scope-based review checklists, review metrics, process anti-patterns
+ - `reference/anti-patterns.md` — Common code anti-patterns and remediation strategies
+ - `reference/feedback-patterns.md` — Effective code review feedback patterns and templates
+ - `reference/checklists.md` — Per-dimension quality checklists for thorough reviews
## Workflow
### 1. Gather Context
-
- Read target code and understand the change context (ticket, requirements). Determine change size and set checklistScope:
-
- match (lineCount) {
- < 100 => Quick
- 100..500 => Standard
- > 500 => Deep
- }
-
- ### 2. Assess Dimensions
-
- Evaluate each dimension systematically:
-
- - **Correctness** — Does it work? Edge cases, error handling, null safety, data validation
- - **Design** — Well-structured? SRP, coupling, cohesion, abstraction, extensibility
- - **Readability** — Understandable? Naming, comments (why not what), complexity (<10 cyclomatic), flow
- - **Security** — Secure? Input validation, auth checks, data exposure, dependency vulnerabilities
- - **Performance** — Efficient? Algorithm complexity, memory, I/O optimization, caching, concurrency
- - **Testability** — Testable? Coverage, behavior-not-implementation tests, mockable deps, determinism
-
- Record strengths alongside findings.
+ - Understand change scope, intent, and affected user/system paths.
- ### 3. Detect Anti-Patterns
+ ### 2. Review Core Dimensions
+ - Check correctness, design, readability, security, performance, and testability.
- Read reference/anti-patterns.md for the detailed catalog.
- Scan for method-level, class-level, and architecture-level anti-patterns.
- Each detected anti-pattern becomes a ReviewFinding with specific remediation.
+ ### 3. Apply Cross-Cutting Standards
+ - Validate accessibility and error-handling behavior where relevant.
### 4. Prioritize Findings
-
- Classify each finding by priority:
-
- match (priority) {
- CRITICAL => Security vulnerabilities, data loss risks, breaking API changes, production stability
- HIGH => Logic errors, performance in hot paths, missing error handling, architectural violations
- MEDIUM => Duplication, missing tests, unclear naming, complex conditionals
- LOW => Style inconsistencies, minor optimizations, documentation, refactoring suggestions
- }
-
- Sort findings:
- 1. Sort by priority descending.
- 2. Break ties by dimension.
-
- ### 5. Format Feedback
-
- Read reference/feedback-patterns.md for tone guidance.
-
- For each finding, apply the feedback formula:
- [Observation] + [Why it matters] + [Suggestion] + [Example if helpful]
-
- Structure output:
- 1. Summary — overall impression, change scope
- 2. Strengths — what's done well (always include)
- 3. Findings table — sorted by priority
- 4. Detailed findings — each with observation, impact, suggestion
- 5. Checklist results — from reference/checklists.md appropriate to scope
+ - Rank by impact and urgency; avoid noisy low-value comments.
+ ### 5. Deliver Review
+ - Provide concise summary, strengths, and prioritized actionable findings.