Credential Audit

Skill: Credential Audit

Purpose

Secrets leak in three ways: hardcoded in source files, committed to git history, or accidentally written to log files. This skill catches all three.

What It Checks

1. Source File Scan

• Scan for patterns that look like API keys, tokens, passwords
• Check all .ts, .js, .json, .yml, .yaml, .sh, .env* files
• Exclude node_modules/, .git/, dist/

2. Git History Scan

• Check recent commits for accidentally committed secrets
• Look for .env files ever committed to git

3. Log File Scan

• Check PM2 logs, startup logs for accidental secret printing
• Look for tokens/keys in log output

4. Permissions Check

• .env.secrets should be 600 (owner read/write only)
• start-system.sh should be 700

Patterns to Detect

sk-[a-zA-Z0-9]{32,}           # OpenAI API key
AIza[0-9A-Za-z_-]{35}         # Google API key
[0-9]{10}:[A-Za-z0-9_-]{35}   # Telegram bot token
op_[a-zA-Z0-9]{32,}           # 1Password service token
password\s*=\s*["'][^"']+["'] # Hardcoded password

Commands

# Scan for common secret patterns
grep -rE "(sk-|AIza|password\s*=|token\s*=|secret\s*=)" \
  /Users/raphael/juliaz_agents \
  --include="*.ts" --include="*.js" --include="*.sh" --include="*.yml" \
  --exclude-dir=node_modules --exclude-dir=.git --exclude-dir=dist \
  -l 2>/dev/null

# Check if .env files were ever committed  
git -C /Users/raphael/juliaz_agents log --all --full-history -- "**/.env*" 2>/dev/null

# Check file permissions
stat -f "%Sp %N" /Users/raphael/juliaz_agents/.env.secrets 2>/dev/null

Severity Rules

Output Format

CREDENTIAL AUDIT
✅ No secrets found in source files
🔴 Telegram token found in orchestrator/logs/pm2-out.log (line 245)
⚠️  .env.secrets permissions: 644 (should be 600)