dirien/yet-another-agent-harness / .claude/skills/security-audit
security-audit
UnverifiedComprehensive security audit covering OWASP Top 10, secrets detection, supply chain security, threat modeling, and language-specific vulnerability patterns. Investigates actual code paths rather than grep-matching keywords. Generates a scored SECURITY_AUDIT.md with prioritized remediation. Use when assessing application security, preparing for a security review, or onboarding to a codebase with security concerns.
| File | Kind | Size |
|---|---|---|
| SKILL.md | skill md | 5.7 KB |
| references/language-security.md | markdown | 3.8 KB |
| references/owasp-patterns.md | markdown | 4.5 KB |
| references/report-template.md | markdown | 3.1 KB |
| references/secrets-and-supply-chain.md | markdown | 2.9 KB |
| references/threat-model.md | markdown | 3.6 KB |
references/threat-model.md
View on GitHub<!-- Copyright 2025-2026 Richard Shade. Licensed under Apache-2.0. -->
<!-- SPDX-License-Identifier: Apache-2.0 -->
# Threat modeling
Apply STRIDE to the actual codebase — not a generic checklist. Map
real entry points, data flows, and trust boundaries.
## Step 1: Identify entry points
Find every way data enters the application:
- **HTTP/API endpoints** — routes, handlers, controllers
- **CLI arguments** — command-line parsers, flag definitions
- **Environment variables** — configuration loaded at startup
- **File system reads** — config files, user uploads, data imports
- **Database queries** — data from external databases
- **Message queues** — event consumers, webhook receivers
- **Inter-service calls** — gRPC, REST, GraphQL clients
For each entry point, note: what data comes in, who can send it,
and what validation exists.
## Step 2: Map trust boundaries
A trust boundary is where data crosses from a less-trusted zone to
a more-trusted zone:
- User input → application logic (most critical)
- External API response → application logic
- Database result → application logic (usually trusted, but verify)
- Application → filesystem (write operations)
- Application → external service (outbound, data leakage risk)
Draw the boundary: everything outside is untrusted, everything
inside is trusted. Validate at every crossing.
## Step 3: Apply STRIDE
For each entry point and trust boundary crossing, evaluate:
**Spoofing** — can an attacker impersonate a legitimate user or
service?
- Check: authentication at this entry point
- Check: certificate validation for service-to-service calls
- Check: token verification (JWT signature, expiry, issuer)
**Tampering** — can an attacker modify data in transit or at rest?
- Check: TLS for all network communication
- Check: integrity checks on file uploads or downloads
- Check: CSRF protection on state-changing operations
- Check: signed cookies/tokens vs unsigned
**Repudiation** — can an attacker perform an action and deny it?
- Check: are security-relevant actions logged?
- Check: are logs tamper-proof (append-only, centralized)?
- Check: is there enough context in logs to reconstruct events?
**Information disclosure** — can an attacker access unauthorized
data?
- Check: error messages — do they leak stack traces, SQL, or
internal paths?
- Check: API responses — do they return more data than needed?
- Check: logging — is sensitive data (passwords, tokens, PII)
written to logs?
- Check: directory listing, source maps, debug endpoints exposed?
**Denial of service** — can an attacker make the system unavailable?
- Check: rate limiting on public endpoints
- Check: request size limits (body, file uploads)
- Check: query complexity limits (GraphQL depth, SQL joins)
- Check: timeout configuration for external calls
- Check: resource exhaustion (memory, connections, file handles)
**Elevation of privilege** — can an attacker gain higher access?
- Check: role-based access control enforced consistently
- Check: privilege escalation via parameter manipulation
- Check: admin functionality accessible without admin role
- Check: insecure direct object references (IDOR)
## Step 4: Prioritize threats
Score each identified threat:
| Factor | Scale |
| ------ | ----- |
| Damage potential | 1-3 (what's the worst case?) |
| Reproducibility | 1-3 (how easy to exploit?) |
| Exploitability | 1-3 (what skill level needed?) |
| Affected users | 1-3 (how many impacted?) |
| Discoverability | 1-3 (how easy to find?) |
Sum = DREAD score (5-15). Map to severity:
- 12-15: CRITICAL
- 9-11: HIGH
- 6-8: MEDIUM
- 5: LOW
Related skills
| Skill | Stars | Updated | Works with | Status | Install |
|---|---|---|---|---|---|
| rea-analysis-optimizationmorluto/reaOptimize REA providers and shared analysis workflows for agent context, latency, CPU, and memory while preserving correctness and Evidence. Use for analysis bottlenecks and cache or query design; CI/test-suite optimization belongs to rea-ci-tests-audit. | Stars: 471k | Updated | Claude CodeCodex | Verified | Install |
| rea-ci-tests-auditmorluto/reaAudit REA CI and test suites for correctness, fixture drift, false passes, and measured performance; use for suite or workflow investigations and efficiency changes, rather than ordinary feature testing or a single test failure. | Stars: 471k | Updated | Claude CodeCodex | Verified | Install |
| rea-tool-designmorluto/reaAudit, investigate, design, or change REA investigation tools, CLI/MCP contracts, provider capabilities, and Evidence semantics. | Stars: 471k | Updated | Claude CodeCodex | Verified | Install |
| ponytail-auditDietrichGebert/ponytailQuality audit of the whole repo: bugs, security, real load, missing tests, speed, and what to delete. Most important first. | Stars: 194k | Updated | Claude CodeCodex | Verified | Install |