Security audit
Security checks for vulnerabilities and agentic risk
This is a small architecture-design guidance skill with some vague activation wording, but no hidden code, persistence, credential handling, or destructive behavior.
Install only if you want an agent to take an architecture-validation posture that may suggest writing prototypes, running commands, tests, scans, or Docker-based checks. Review proposed changes and commands before letting it modify a real project or infrastructure.
The description says to use the skill when requests 'match this domain or role,' but it does not define concrete trigger phrases, scope boundaries, or exclusion conditions. This broad wording could overlap with many general software-design requests and lead to unintended invocation.
The instruction to use the skill when work matches the 'architecture-designer' specialist role is circular and does not clarify what qualifies or does not qualify as that role. Without concrete constraints or examples, the trigger condition remains ambiguous.
The skill explicitly enumerates an internal local skill path (~/.claude/skills/architecture-patterns/SKILL.md), which discloses implementation details about the agent environment and available capability structure. In adversarial settings, this kind of internal path and skill discovery information can aid prompt-targeting, environment fingerprinting, or chaining attacks against adjacent skills.
## Skill Invocation
**For pattern selection and API design:** `~/.claude/skills/architecture-patterns/SKILL.md`
**Skill provides:**
- Application patterns (Monolith, Modular Monolith, Microservices) + decision trees
No suspicious patterns detected.