perfection-audit
Skill2 starsChanged 8 months agoMITView on GitHub
Full perfection audit: security, performance, code quality, type safety - Enterprise standard
- Reads credentials
In this repository34 other files it gives its agents.
- global/CLAUDE.mdCLAUDE.md
- templates/CLAUDE.mdCLAUDE.md
- accessibility-reviewSkill
- api-designSkill
- audit-runnerSkill
- auto-fixSkill
- brainstormSkill
- compliance-checkSkill
- continuous-testing-guardianSkill
- contract-lookupSkill
- explainSkill
- idea-validatorSkill
- load-test-auditSkill
- monthly-reviewSkill
- new-featureSkill
- nextjs-reviewSkill
- performance-deep-auditSkill
- post-code-security-auditSkill
- pre-code-security-checkSkill
- pre-deploymentSkill
- project-setupSkill
- pro-samuelSkill
- quick-auditSkill
- risk-scoreSkill
- roadmap-builderSkill
- secret-scannerSkill
- seo-auditSkill
- supabase-reviewSkill
- token-economySkill
- typescript-reviewSkill
- ux-cognitive-load-checkSkill
- ux-heuristic-auditSkill
- ux-progressive-disclosureSkill
- welcomeSkill
The file
We split this skill into the rules it gives an agent
2 rules. About 0 are common, 1 is distinctive, and 1 is very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈1
Distinctive · estimate
This author's own choices. Worth a look.
1
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Very specific: line 15
Verify that a Next.js/Supabase application reaches perfection standards on:
- Distinctive: line 528
Run manually before each release
≈ “Common” is an estimate: at least 1 in 5 of the 364,980 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
🏆 Perfection Audit - Enterprise Grade
Exhaustive audit based on the standards of the best tech companies (Google, Vercel, AWS) and security firms (Cure53, Trail of Bits, NCC Group).
Objective
Verify that a Next.js/Supabase application reaches perfection standards on:
- ✅ Security (OWASP Top 10, ASVS Level 2)
- ✅ Performance (Core Web Vitals, load testing)
- ✅ Code quality (TypeScript strict, zero console, centralized logger)
- ✅ Type safety (zero
any, complete interfaces) - ✅ Scalability (100+ concurrent users)
Phase 1: TypeScript Health (5 min)
Execution
npx tsc --noEmit --prettyMetrics
- Target: 0 errors
- Acceptable: < 5 non-critical errors
- Blocking: > 10 errors OR 1+ critical error
Check
- 0 TypeScript errors
strict: trueintsconfig.jsonnoImplicitAny: truestrictNullChecks: true- No
@ts-ignoreexcept justified cases
Phase 2: Critical Security (20 min)
2.1 API Keys & Secrets
Automatic search:
# Keys exposed to the client
grep -r "NEXT_PUBLIC_.*KEY\|NEXT_PUBLIC_.*SECRET" .env* app/ components/
# Service role keys client-side
grep -r "SERVICE_ROLE_KEY" app/ components/ lib/Checklist:
- Zero
NEXT_PUBLIC_for API keys (except Maps, Analytics) SUPABASE_SERVICE_ROLE_KEYserver-side only- Stripe secret key never exposed
.envin.gitignore
2.2 Debug Endpoints
Search:
find app/api/debug -type f
grep -r "/api/test\|/api/debug" app/api/Checklist:
- Zero
/api/debug/endpoints - Zero
/api/test/endpoints - Admin endpoints protected by an
is_admincheck
2.3 Rate Limiting
Search:
# Public routes without rate limit
find app/api -name "route.ts" -exec grep -L "ratelimit\|RateLimit" {} \;Checklist:
- All public routes rate-limited
- Login: 5/min per IP
- Signup: 3/hour per IP
- Upload: 10/hour per user
- Password reset: 3/hour per email
2.4 RLS (Row Level Security)
SQL to run:
-- Tables without RLS
SELECT tablename
FROM pg_tables
WHERE schemaname = 'public'
AND rowsecurity = false;
-- Suspicious policies
SELECT tablename, policyname, definition
FROM pg_policies
WHERE definition NOT LIKE '%auth.uid()%';Checklist:
- RLS enabled on all sensitive tables
- Policies tested with different users
- Service role usage justified (webhooks, cron)
2.5 Authentication
Search:
# API routes without auth
find app/api -name "route.ts" -exec grep -L "session\|getUser\|auth" {} \;Checklist:
- Auth check on all protected routes
- Session timeout configured and ENFORCED
- Password hashing bcrypt/Argon2 (NOT SHA256)
- MFA available for critical accounts
Phase 3: Performance & Scalability (15 min)
3.1 Database Query Optimization
Anti-pattern search:
# SELECT * instead of specific columns
grep -r "select('\*')" app/ lib/ --include="*.ts"
# Queries without .limit()
grep -r "\.from(" app/ lib/ | grep -v "\.limit(" | grep -v "\.single()"
# N+1 queries (loop with query)
grep -rA 10 "for.*of\|\.map(" lib/services/ | grep "supabase.from"Checklist:
- Zero
SELECT *count queries .limit()on all list queries- Zero N+1 queries
- Indexes created on filtered columns
3.2 Image Optimization
Search:
# <img> instead of <Image>
grep -r "<img" app/ components/ --include="*.tsx"Checklist:
- Next.js
<Image>everywhere (except SVG) - Lazy loading enabled
next.config.jsdomains configured- WebP/AVIF formats enabled
3.3 React Performance
Search:
# Components without memo
grep -r "export.*function.*Props" components/ | grep -v "memo"
# Inline callbacks in props
grep -r "onClick={() =>" app/ components/Checklist:
- Frequently rendered components wrapped in
memo() - Callbacks wrapped in
useCallback - Expensive calculations in
useMemo - Context providers memoized
3.4 Load Testing Readiness
Questions:
- How many simultaneous connections can Supabase handle?
- How many requests/second do the API routes support?
- Is the Vercel CDN enabled?
Checklist:
- Supabase connection pooling configured
- API routes stateless (horizontally scalable)
- Correct cache headers (
Cache-Control) - Rate limiting prevents DoS
Benchmark targets:
- 100 simultaneous users: < 500ms p95
- 1000 req/sec: API routes must hold
- Core Web Vitals: LCP < 2.5s, FID < 100ms, CLS < 0.1
Phase 4: Code Quality (10 min)
4.1 Console Statements
Search:
grep -r "console\.\(log\|error\|warn\)" app/ lib/ components/ contexts/ \
--include="*.ts" --include="*.tsx" \
| grep -v "logger.ts" \
| wc -lTarget: 0 console statements in production
Checklist:
- Zero
console.log - Zero
console.error - Centralized logger used everywhere
- Dev-only logs wrapped in
if (process.env.NODE_ENV === 'development')
4.2 Type Safety
Search:
# as any
grep -r "as any" app/ lib/ components/ --include="*.ts" --include="*.tsx" | wc -l
# @ts-ignore
grep -r "@ts-ignore\|@ts-expect-error" app/ lib/ components/Checklist:
- < 50
as any(none in critical code) - < 5
@ts-ignore(all justified) - Interfaces for all data structures
- Type guards for unions
4.3 Error Handling
Search:
# Silent catch
grep -rA 3 "} catch" app/ lib/ | grep -B 3 "^$"
# Error boundaries
find app -name "error.tsx"Checklist:
- Error boundary on each segment (owner, seeker, resident, etc.)
- All catch blocks log via the logger
- User-friendly UI errors
- Stack traces never exposed
4.4 Dead Code
Search:
# Archived files
find app lib components -name "*OLD*" -o -name "*BACKUP*" -o -name "_archived"
# Unused imports
npx eslint . --ext .ts,.tsx --rule 'no-unused-vars: error' 2>&1 | grep "is defined but never used"Checklist:
- Zero
*-OLD.tsxfiles - Zero
_archived/folders - Imports cleaned up (ESLint)
Phase 5: Supabase Specifics (10 min)
5.1 Connection Pooling
Check DATABASE_URL:
postgresql://[user]:[password]@[host]:6543/postgres?pgbouncer=true&connection_limit=10Supabase limits:
- Free tier: 60 connections max
- Pro tier: 200 connections max
- Enterprise: unlimited
For 100 simultaneous users:
- Connection pooling REQUIRED
pgbouncer=truein the connection stringconnection_limit=10per instance
5.2 Realtime Limits
Free tier:
- 200 concurrent connections
- 2 concurrent channels
If > 200 simultaneous users:
- Upgrade to Pro ($25/month)
- Or disable non-essential realtime
5.3 Storage Limits
Free tier:
- 1 GB storage
- 2 GB bandwidth/month
If the app handles photos:
- Monitor current usage
- Plan an upgrade or aggressive compression
Phase 6: Build & Deployment (5 min)
6.1 Build Verification
npm run buildChecklist:
- Build passes without errors
- Zero TypeScript warnings
- Reasonable bundle size (< 500KB First Load JS)
- Source maps disabled in production
6.2 Vercel Configuration
Required vercel.json:
{
"headers": [
{
"source": "/(.*)",
"headers": [
{ "key": "X-Frame-Options", "value": "DENY" },
{ "key": "X-Content-Type-Options", "value": "nosniff" },
{ "key": "Strict-Transport-Security", "value": "max-age=63072000" }
]
}
],
"env": {
"NODE_ENV": "production"
}
}Checklist:
- Security headers configured
- Environment variables defined on Vercel
- Custom domain with HTTPS
- Edge config if needed (rate limiting, feature flags)
Phase 7: Monitoring & Observability (5 min)
7.1 Error Tracking
Checklist:
- Sentry configured and working
- Source maps uploaded (Vercel + Sentry)
- Error boundaries capture and log
- Alerts configured (Slack, email)
7.2 Performance Monitoring
Core Web Vitals:
- LCP < 2.5s (75th percentile)
- FID < 100ms
- CLS < 0.1
API Performance:
- p95 < 500ms on critical routes
- Timeouts configured (10s max)
- Slow query logging
Phase 8: Compliance & Legal (5 min)
GDPR (if EU users)
Checklist:
- Cookie consent banner
- Privacy policy accessible
- User data export available
- Right to be forgotten implemented
- Data retention policy documented
PCI DSS (if payments)
Checklist:
- NO credit card storage
- Stripe Checkout OR Elements (PCI compliant)
- HTTPS mandatory
- Logs do NOT contain card numbers
Audit Report - Template
# PERFECTION AUDIT - [Project Name]
**Date**: [YYYY-MM-DD]
**Duration**: [Xh]
**Auditor**: Claude Opus 4.6 Perfection Audit
---
## 📊 OVERALL RESULTS
| Phase | Score | Status |
|-------|-------|--------|
| TypeScript | X/100 | [✅ PASS / 🔴 FAIL] |
| Security | X/100 | [✅ PASS / 🔴 FAIL] |
| Performance | X/100 | [✅ PASS / 🔴 FAIL] |
| Code Quality | X/100 | [✅ PASS / 🔴 FAIL] |
| Scalability | X/100 | [✅ PASS / 🔴 FAIL] |
**Overall score**: XX/100
**Verdict**:
- ≥ 95%: 🏆 PERFECTION - Production ready
- 85-94%: ✅ EXCELLENT - Minor improvements
- 75-84%: 🟡 GOOD - Fix non-critical issues
- < 75%: 🔴 INSUFFICIENT - Blocking issues
---
## 🔴 CRITICAL FINDINGS
### [CRIT-001] [Title]
- **Severity**: CRITICAL
- **Location**: `[file:line]`
- **Issue**: [Description]
- **Impact**: [Security/Performance consequence]
- **Fix**: [Action required]
- **Effort**: [Xh]
---
## 🟠 HIGH PRIORITY
[List of high priority issues]
---
## 🟡 MEDIUM PRIORITY
[List of medium priority issues]
---
## ✅ PASSED CHECKS
- TypeScript: 0 errors
- console.log: 0 statements
- API keys: Secured
- [etc.]
---
## 📈 RECOMMENDATIONS
### Immediate (P0)
1. [Action]
### Short-term (P1)
1. [Action]
### Long-term
1. [Action]
---
## 📊 METRICS SUMMARY
TypeScript errors: 0 console.log: 0 console.error: 0 as any: XX Critical security: 0 issues Rate limiting: 100% Error boundaries: 6 Logger usage: XXX occurrences Build: CLEAN
---
## 🔄 NEXT STEPS
1. Fix CRITICAL findings
2. Address HIGH priority
3. Schedule follow-up audit in 3 months
4. Update security patterns documentUsage
Full audit
/perfection-auditTargeted audit (security only)
/perfection-audit --security-onlyAudit with auto-fix
/perfection-audit --auto-fixContinuous Integration
This skill can be:
- Run manually before each release
- Integrated into CI/CD (GitHub Actions)
- Triggered automatically after a major refactoring
- Used for quarterly audits
Sources & Standards
- OWASP Top 10 2025
- ASVS Level 2
- Google Cloud Security Best Practices
- Vercel Production Checklist
- Next.js Security Guide
- Supabase Security Practices
- Core Web Vitals Standards
- CWE Top 25
Updated: 2026-02-13 Based on: a real audit (65 errors → 0, 2663 console → 0)
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

