Security audit
Security checks for vulnerabilities and agentic risk
This is a read-only code review skill with disclosed optional cross-model review and no evidence of hidden execution, persistence, or data exfiltration.
Install this if you want stricter code-review feedback. Be aware that the optional --dual mode may use an additional model and more time or cost, and its tone is intentionally direct.
The description lists generic phrases such as "find issues," "what's wrong with this code," and "review this code" as activation examples. In a markdown skill file, these are broad enough to match ordinary conversation and there are no explicit exclusion conditions or tighter scope constraints to prevent accidental invocation.
Skill allows unbounded resource consumption (API calls, storage, compute). Without rate limits or quotas, a compromised or misbehaving agent can cause denial-of-service or cost overruns.
- N+1 query patterns
- Raw string interpolation in queries (SQL injection risk)
- Missing indexes on frequently queried columns
- Unbounded queries without LIMIT
## Operating Constraints
No suspicious patterns detected.