Skip to content
Back to skills

Code Analyze

ASecurity

Analyze codebase for patterns, issues, and improvements. Enforces analysis standards while incorporating user-specific requirements.

  • 666 stars
  • 0 votes
  • 0 copies
  • 4 views
  • Added February 8, 2026
code-qualitybashsqlnodetestingrefactoringgitapidatabasesecurityperformance

Works with

  • api

Security analysis

A100/100

Scanned February 12, 2026

npx -y skills add majiayu000/claude-skill-registry --skill code-analyze --agent claude-code

Installs into .claude/skills of the current project.

Are you the author of Code Analyze?

Add the live security badge to your README. It updates with every re-scan.

Security grade badge for Code Analyze
[![Security: A — Skills Directory](https://www.skillsdirectory.com/api/skills/majiayu000-code-analyze/badge)](https://www.skillsdirectory.com/skills/majiayu000-code-analyze)

More formats (shields.io, HTML) on the badges page. Keep it an A: scan every change in CI with Pro.

Download with Pro
SKILL.md
---
name: code-analyze
description: Analyze codebase for patterns, issues, and improvements. Enforces analysis standards while incorporating user-specific requirements.
---

# Code Analyzer Skill

## What This Skill Does

Performs standardized code analysis with user context:
- Analyzes codebase structure and patterns
- Identifies technical debt and issues
- Suggests improvements and refactoring
- Enforces coding standards
- Generates actionable reports

## When Claude Should Use This

Use this skill when the user:
- Says "analyze the code" or "review the codebase"
- Wants to find bugs or issues
- Needs architecture review
- Mentions code quality or technical debt
- Requests security or performance analysis

## Analysis Workflow

### 1. Gather User Context
```
Ask the user:
1. Focus areas? (security, performance, architecture, all)
2. Specific concerns?
3. Depth of analysis? (quick scan, standard, deep dive)
4. Output format? (report, tasks, PR comments)
```

### 2. Standardized Analysis Process

#### Phase 1: Structure Analysis
```bash
# Map codebase structure
find . -type f -name "*.ts" -o -name "*.js" | head -20

# Count lines of code
cloc . --exclude-dir=node_modules,dist,out

# Identify entry points
grep -r "export.*function\|export.*class" --include="*.ts" | head -10
```

#### Phase 2: Pattern Detection
```bash
# Find potential issues
grep -r "TODO\|FIXME\|HACK\|XXX" --include="*.ts"

# Identify deprecated patterns
grep -r "deprecated\|@deprecated" --include="*.ts"

# Find console.logs (potential debug code)
grep -r "console\." --include="*.ts" --exclude-dir=node_modules
```

#### Phase 3: Dependency Analysis
```bash
# Check for outdated packages
npm outdated

# Audit for vulnerabilities
npm audit

# Analyze bundle size
npm ls --depth=0
```

### 3. Automated Report Generation

Generate `analysis/ANALYSIS_[DATE].md`:
```markdown
# Code Analysis Report

## Summary
- **Date**: YYYY-MM-DD
- **Scope**: [User specified focus]
- **Health Score**: X/100

## Critical Issues
1. [Issue]: [Description]
   - File: [path:line]
   - Severity: High/Medium/Low
   - Fix: [Recommendation]

## Architecture Review
- **Patterns Found**: [List]
- **Anti-patterns**: [List]
- **Suggestions**: [List]

## Technical Debt
- **Total**: X hours estimated
- **Priority Items**: [List]

## Action Items
- [ ] Fix critical security issues
- [ ] Refactor deprecated patterns
- [ ] Update dependencies
```

### 4. Integration with Git Workflow

#### For Issues Found:
```bash
# Create issue branch
git checkout -b fix/[issue-name]

# For refactoring needs
git checkout -b refactor/[component-name]

# For security fixes
git checkout -b security/[vulnerability-name]
```

### 5. Enforcement Rules

#### Must Fix (Blocking):
- Security vulnerabilities (High/Critical)
- Breaking changes in dependencies
- Memory leaks
- Exposed secrets/credentials

#### Should Fix (Warning):
- Deprecated API usage
- Performance bottlenecks
- Code duplication > 50 lines
- Circular dependencies

#### Consider Fixing (Info):
- TODO/FIXME comments
- Missing documentation
- Complex functions (cyclomatic > 10)
- Long files (> 500 lines)

## Analysis Types

### 1. Security Analysis
```bash
# Check for secrets
grep -r "api[_-]key\|password\|secret\|token" --include="*.ts"

# Review authentication
grep -r "authenticate\|authorize\|jwt\|session" --include="*.ts"

# Check HTTPS usage
grep -r "http://" --include="*.ts"
```

### 2. Performance Analysis
```bash
# Find potential bottlenecks
grep -r "for.*for\|while.*while" --include="*.ts"

# Check for synchronous operations
grep -r "readFileSync\|execSync" --include="*.ts"

# Identify heavy operations
grep -r "sort\|filter.*map\|reduce" --include="*.ts"
```

### 3. Architecture Analysis
- Component coupling
- Layer violations
- Dependency cycles
- Interface segregation
- Single responsibility

## Output Actions

### 1. Create Fix Branches
For each critical issue:
```bash
git checkout -b fix/[issue-id]
echo "Fix plan" > fixes/[issue-id].md
```

### 2. Generate Tasks
Create `tasks/FIXES_[DATE].md`:
```markdown
## Critical Fixes Required

### HIGH Priority
- [ ] Fix SQL injection in [file:line]
- [ ] Remove hardcoded credentials in [file:line]

### MEDIUM Priority
- [ ] Refactor [component] to reduce complexity
- [ ] Update deprecated [package] usage

### LOW Priority
- [ ] Add missing documentation
- [ ] Clean up TODO comments
```

### 3. PR Templates
Generate `.github/PULL_REQUEST_TEMPLATE/fix.md`:
```markdown
## Fix for: [Issue ID]

### What was wrong?
[Description]

### How was it fixed?
[Approach]

### Testing
- [ ] Unit tests pass
- [ ] Integration tests pass
- [ ] Security scan clean
```

## Integration with User Intent

The analyzer combines:
1. **User's specific concerns** (from prompt)
2. **Standardized checks** (always run)
3. **Codebase context** (patterns, history)
4. **Best practices** (industry standards)

Example:
- User: "Check our API for security issues"
- Analyzer: Runs standard checks + deep API security scan

## Protection Against Breaking Changes

### Allowed Fixes:
- Security patches that don't change interfaces
- Performance improvements with same behavior
- Bug fixes with tests
- Documentation updates

### Requires Approval:
- API changes
- Database schema changes
- Configuration changes
- Dependency major version updates

### Never Auto-Fix:
- Core business logic
- Authentication/Authorization
- Payment processing
- Data migrations

## Success Metrics

Analysis succeeds when:
- [ ] All critical issues identified
- [ ] Report generated with actionable items
- [ ] Fix branches created for issues
- [ ] Team understands findings
- [ ] Clear path to resolution

Attribution

Is this your skill, or is something wrong with this listing? Request removal or report an issue. Author removals are honored within 72 hours.

Comments

Loading comments…

Newsletter

Get the weekly list of safe new skills, and the ones we pulled.