security-audit

---

name: security-audit description: SDLC Phase: ANALYSIS. Dedicated security and privacy audit — OWASP Top 10, PII handling, authentication, authorization, session security, rate limiting, secrets management, and dependency vulnerabilities. upstream:

• audit-protocol downstream:
• roadmap-planning
• code-refactoring

---

Security Audit

Prerequisite: Load context-awareness for project orientation. Running audit-protocol first provides the baseline code quality audit.

When to Activate

Use this skill for a dedicated security and privacy audit. Covers OWASP Top 10, PII handling, authentication, authorization, session security, rate limiting, secrets management, and dependency vulnerabilities.

Agent Workflow

Using this skill follows 4 phases:

1. Construct — Knowledge, Context & Scope

• Load context-awareness skill for project orientation
• Read relevant docs: module docs, pattern docs, reference docs
• Understand task scope: what needs to be done, which files are affected
• Verify paths, class names, signatures against actual code (don't trust docs blindly)
• Determine approach: at least 2 options before deciding

2. Execute — Security Audit Execution

• Audit authentication: password hashing, rate limiting, recovery flows
• Audit authorization: Policy methods, super admin bypass, permissions config
• Audit XSS: Blade escaping, {!! !!} occurrences, CSP headers
• Audit SQL injection: parameterized binding, raw SQL check
• Audit mass assignment: #[Fillable], no $request->all()
• Audit PII: data isolation, log masking, GDPR deletion path
• Output: GitHub Issues with security vulnerability reports including severity, impact, and fix recommendations

3. Verify — Quality Gates

• Run linter: vendor/bin/pint --dirty --format agent
• Run static analysis: vendor/bin/phpstan analyse --no-progress
• Run unit/feature tests: php artisan test --compact --filter={TestName}
• Ensure pre-commit checklist is satisfied
• Check no debug calls (dd/dump/ray) were left behind

4. Report & Commit

• Deliver a comprehensive report to the user:
  • Summary of security findings by category
  • Severity distribution (critical/high/medium/low)
  • Vulnerabilities confirmed closed
• Feeds into: roadmap-planning (prioritize security fixes), code-refactoring (fix vulnerabilities)
• Commit using format: type(scope): description
• Push if requested

Phase Context

Role	Skill
Upstream	audit-protocol (baseline audit)
This skill	ANALYSIS — security-specific
Downstream	roadmap-planning (prioritize fixes), code-refactoring (fix issues)

Audit Categories

Authentication

• Password hashing uses bcrypt/argon2 via Laravel defaults
• Login rate limiting applied (check bootstrap/app.php)
• Account recovery rate limited (recovery slip, password reset)
• Session management follows Laravel best practices
• MFA readiness (future)

Authorization

• Super admin bypass via Gate::before — verify it exists and is not removable
• Policy methods return boolean — check every method
• No inline Gate::authorize() bypassing Policy layer
• Permissions defined in config/permission.php — no magic strings in code
• Check that 5 flat roles are enforced (no role inheritance)

XSS Prevention

• All Blade output uses {{ }} (double curly braces) — escaped
• Every {!! !!} (unescaped) has an inline justification comment
• No inline <script> tags — everything uses Alpine.js
• CSP enforced via SecurityHeaders middleware
• Check CSP allows only necessary external resources

SQL Injection

• No whereRaw() / DB::raw() without parameterized binding
• No string concatenation in query builder
• where('column', $value) used over whereRaw("column = '$value'")

Mass Assignment

• Every model uses #[Fillable] attribute — not $fillable or $guarded
• No Model::create($request->all()) anywhere
• No Model::create($this->all()) in Livewire

File Upload Security

• ALL uploads go through Spatie MediaLibrary
• MIME type validated server-side (not just extension)
• Filenames sanitized with Str::slug()
• File size limits defined per collection

PII & Data Protection

• User profiles stored in separate table from credentials
• Check app/Core/Support/PiiMasker.php — PII masking in logs
• Activity log does not store raw PII
• GDPR deletion path exists (gdpr_deletion_logs table)

Secrets & Configuration

• No hardcoded secrets in code or config files
• .env excluded from version control (check .gitignore)
• APP_KEY must be unique per installation
• Database credentials in .env only

Dependencies

• Check composer.json for known vulnerabilities
• Verify package versions are current
• Check dependabot.yml for automated scanning

Cross-Cutting

• CSRF: all state-changing HTML forms include @csrf or use Livewire
• Rate limiting on: login, password reset, recovery slip, setup token
• Session timeout configured appropriately
• HTTPS enforced in production

Issue Format

Each finding should include:

• Category: Which audit category
• Location: File path and line number
• Vulnerability: OWASP category or specific risk
• Severity: Critical / High / Medium / Low
• Reproduction: Steps if applicable
• Fix: Recommendation

Key Rules

1. Verify each finding manually — automated scans produce false positives
2. Record all findings even if out of scope — prioritization happens downstream
3. Do NOT fix during audit — separate concerns
4. Check existing issues before filing duplicates

References

Topic	Doc
Security conventions	docs/conventions.md (§3)
RBAC & authentication	docs/foundation/rbac.md
Account recovery	docs/foundation/account-recovery.md
Exception hierarchy	docs/architecture/exception-pattern.md
CSP & middleware	app/Core/Http/Middleware/SecurityHeaders.php
File upload security	docs/infrastructure/media-library.md
Session configuration	docs/infrastructure/session.md
Rate limiting	bootstrap/app.php