Skip to content
Development
Command

/review

Comprehensive code review with context-aware analysis and architectural insights

From plugin
claude-cmd
313180 skills180 commands

How it fires

How this command gets triggered: by you, by Claude, or both.

  • Fires itselfClaude auto-loads it when your prompt matches the work.
  • You can call itInvoke it directly when you want it.
  • Slash command/review

Context preview

What this command does when you run it.

Comprehensive code review with context-aware analysis and architectural insights

Command definition

review.md
allowed-tools: Read, Grep, Task, Bash(rg:*), Bash(fd:*), Bash(git:*), Bash(deno:*), Bash(cargo:*), Bash(go:*), Bash(npm:*)
name: "Review"
description: "Comprehensive code review with context-aware analysis and architectural insights"
author: "wcygan"
tags: ["tool"]
version: "1.0.0"
created_at: "2025-07-14T00:00:00Z"
updated_at: "2025-07-14T00:00:00Z"

/review

Context

  • Session ID: !`gdate +%s%N || date +%s%N`
  • Project type: !`fd -t f 'package.json|deno.json|Cargo.toml|go.mod|pom.xml' -d 1 | head -1 | xargs basename || echo 'unknown'`
  • Git status: !`git status --porcelain | head -10 || echo 'Not a git repository'`
  • Recent changes: !`git log --oneline -5 --since='1 week ago' || echo 'No recent commits'`
  • Current branch: !`git branch --show-current || echo 'No current branch'`
  • Build status: !`test -f deno.json && deno check . 2>&1 | tail -3 || test -f Cargo.toml && cargo check --quiet 2>&1 | tail -3 || test -f package.json && npm run build --silent 2>&1 | tail -3 || echo 'No build system detected'`
  • Test status: !`test -f deno.json && deno test --quiet 2>&1 | tail -1 || test -f Cargo.toml && cargo test --quiet 2>&1 | tail -1 || test -f package.json && npm test 2>&1 | tail -1 || echo 'No test runner detected'`
  • File count: !`fd -t f . | wc -l || find . -type f | wc -l`
  • Code files: !`fd -e ts -e js -e rs -e go -e java -e py | wc -l || find . -name '*.ts' -o -name '*.js' -o -name '*.rs' -o -name '*.go' -o -name '*.java' -o -name '*.py' | wc -l`

Your task

STEP 1: Initialize review session

  • Load target files from $ARGUMENTS or analyze entire project if no specific targets
  • Create session state file: /tmp/review-state-$SESSION_ID.json
  • IF analyzing large codebase (>50 files):

**CRITICAL: Deploy parallel sub-agents for maximum performance (10x faster comprehensive review)**

IMMEDIATELY launch 8 specialized parallel review agents:

  • **Agent 1: Architecture & Design Patterns**: Analyze system architecture, design patterns, and structural integrity
  • Focus: Component relationships, abstraction layers, separation of concerns, architectural violations
  • Tools: Dependency analysis, pattern recognition, architectural metrics
  • Expected: Architectural assessment with improvement recommendations
  • **Agent 2: Security Vulnerability Assessment**: Comprehensive security analysis across all attack vectors
  • Focus: Input validation, authentication, authorization, data exposure, injection vulnerabilities
  • Tools: Security pattern analysis, vulnerability scanning, OWASP compliance
  • Expected: Security risk assessment with prioritized remediation plan
  • **Agent 3: Performance & Optimization Review**: Identify performance bottlenecks and optimization opportunities
  • Focus: Algorithm efficiency, database queries, memory usage, caching strategies
  • Tools: Complexity analysis, profiling patterns, resource utilization assessment
  • Expected: Performance optimization roadmap with impact estimates
  • **Agent 4: Testing Coverage & Quality Analysis**: Evaluate test completeness, quality, and maintainability
  • Focus: Test coverage gaps, test quality metrics, testing strategies, mock usage
  • Tools: Coverage analysis, test pattern evaluation, assertion quality assessment
  • Expected: Testing improvement plan with coverage targets
  • **Agent 5: Dependencies & Technical Debt Evaluation**: Assess dependency health and technical debt accumulation
  • Focus: Dependency vulnerabilities, version currency, license compliance, debt patterns
  • Tools: Dependency analysis, debt quantification, update impact assessment
  • Expected: Dependency management strategy with debt reduction plan
  • **Agent 6: Code Quality & Maintainability**: Analyze code quality metrics and maintainability factors
  • Focus: Code complexity, duplication, readability, documentation completeness
  • Tools: Quality metrics calculation, complexity analysis, documentation assessment
  • Expected: Code quality improvement recommendations with refactoring priorities
  • **Agent 7: API & Interface Design**: Review API design, interface consistency, and integration patterns
  • Focus: API contracts, versioning, backwards compatibility, interface design principles
  • Tools: API analysis, contract validation, integration pattern assessment
  • Expected: API design recommendations with versioning strategy
  • **Agent 8: Configuration & DevOps Review**: Analyze build, deployment, and operational configurations
  • Focus: Build scripts, CI/CD pipelines, environment configurations, operational readiness
  • Tools: Configuration analysis, pipeline assessment, operational pattern review
  • Expected: DevOps optimization recommendations with deployment improvements

**Sub-Agent Coordination:**

  • Each agent saves findings to `/tmp/review-agents-$SESSION_ID/`
  • Parallel execution provides 10x speed improvement over sequential review
  • Cross-agent correlation identifies systemic issues across multiple domains
  • Results synthesized into comprehensive review report with actionable priorities
  • ELSE:
  • Proceed with focused single-agent comprehensive analysis

STEP 2: Context gathering

  • Load project configuration files automatically:
  • @package.json @deno.json @Cargo.toml @go.mod @pom.xml
  • @tsconfig.json @*.d.ts
  • @README.md @CLAUDE.md @CHANGELOG.md
  • @docker-compose.yml @.env.example
  • Discover test files: @**/_.test._ @**/_.spec._
  • Identify configuration and infrastructure files
  • Map project structure and dependencies

STEP 3: Multi-dimensional analysis

FOR EACH target file or component:

**Architecture Analysis:**

  • Design patterns and architectural decisions
  • Separation of concerns and modularity
  • Code organization and structure
  • Interface design and abstractions

**Security Assessment:**

  • Input validation and sanitization
  • Authentication and authorization
  • Data handling and storage
  • Potential vulnerabilities and attack vectors

**Performance Evaluation:**

  • Algorithm efficiency and complexity
Read more
Ships withclaude-cmd

A lightweight (~46kB) and comprehensive CLI tool for managing Claude commands, configurations, and workflows.

Get the whole plugin