ai-tasks
Pass
Audited by Gen Agent Trust Hub on Jul 2, 2026
Risk Level: SAFEPROMPT_INJECTION
Full Analysis
- [SAFE]: No malicious patterns, obfuscation, or dangerous command executions were detected in the skill instructions or examples.
- [PROMPT_INJECTION]: Indirect Prompt Injection Surface. 1. Ingestion points: Untrusted visual content (images/videos) referenced in SKILL.md. 2. Boundary markers: Absent; no delimiters are defined to separate image content from metadata instructions. 3. Capability inventory: Capability to modify asset tags and custom metadata fields in the Digital Asset Management (DAM) system as shown in resources/EXAMPLES.md. 4. Sanitization: Absent; the skill does not specify validation or escaping for the AI-extracted metadata.
Audit Metadata