injection-check
Pass
Audited by Gen Agent Trust Hub on Sep 14, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill consists of instructional text for identifying injection vulnerabilities such as SQLi, Command Injection, and XSS. No malicious patterns, obfuscation, or data exfiltration attempts were found.
- [INDIRECT_PROMPT_INJECTION]: The skill reviews code construction logic that may incorporate untrusted data, presenting a surface for indirect prompt injection. This risk is mitigated by the skill's lack of executable tools or write capabilities. Evidence: 1. Ingestion points: External code provided to the agent for review. 2. Boundary markers: Absent. 3. Capability inventory: None (text-only diagnostics). 4. Sanitization: Absent.
Audit Metadata