figure-designer
Pass
Audited by Gen Agent Trust Hub on Sep 15, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process untrusted external data, specifically user-provided text descriptions and image files, which creates a surface for indirect prompt injection.
- Ingestion points: The skill takes the user's intent (storytelling need), research context (area, venue), and an image file path (via the 'Read' tool) as defined in
SKILL.md. - Boundary markers: The instructions do not define explicit delimiters (e.g., XML tags) or specific warnings to ignore instructions embedded within the processed image or text context.
- Capability inventory: The skill utilizes a 'Read' tool to inspect local images. Analysis of the scripts and instructions reveals no additional high-risk capabilities such as network exfiltration, arbitrary command execution (exec/eval), or unauthorized file-write operations.
- Sanitization: There is no evidence of sanitization, validation, or escaping logic applied to the external content before it is processed by the agent.
Audit Metadata