media-accessibility
Pass
Audited by Gen Agent Trust Hub on Oct 31, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill instructions are limited to accessibility auditing of media elements.
- [SAFE]: No network operations, sensitive file access, or command execution patterns were detected.
- [SAFE]: The skill configuration uses disable-model-invocation: true, which restricts model behavior.
- [INDIRECT_PROMPT_INJECTION]: Analysis of vulnerability surface: 1. Ingestion points: , , elements and caption files (SKILL.md). 2. Boundary markers: Absent. 3. Capability inventory: None (no tools or dangerous commands). 4. Sanitization: Absent. The lack of exploitable capabilities makes this surface safe.
Audit Metadata