agent-patterns

Pass

Audited by Gen Agent Trust Hub on Mar 13, 2026

Risk Level: SAFENO_CODE
Full Analysis
  • [SAFE]: No malicious patterns, obfuscation, or unauthorized external communications were detected. The skill is entirely composed of documentation files.
  • [NO_CODE]: The skill consists exclusively of markdown documentation and lacks any executable scripts or binary files, limiting its risk to information retrieval.
  • [PROMPT_INJECTION]: The skill possesses a surface for indirect prompt injection as it processes user-triggered keywords to read local documentation. * Ingestion points: SKILL.md (triggers), references/ui-test-report.md * Boundary markers: Absent * Capability inventory: Read tool only * Sanitization: Absent
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 13, 2026, 06:58 AM