diagram-design-editorial
Pass
Audited by Gen Agent Trust Hub on Sep 14, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The 'Brand Onboarding' feature instructs the agent to fetch and process content from an untrusted external URL provided by the user. This creates a surface where malicious instructions embedded in the external website's HTML could potentially influence the agent's behavior during the style extraction and configuration process.
- Ingestion points: External URLs provided in the onboarding command (referenced in SKILL.md).
- Boundary markers: The instructions do not define clear boundaries or isolation markers to distinguish the fetched untrusted content from the agent's own instructions.
- Capability inventory: The skill utilizes the agent's ability to fetch remote content via URL and write to local configuration files (references/style-guide.md).
- Sanitization: There are no explicit instructions for sanitizing, validating, or filtering the fetched content before it is processed by the agent.
- [EXTERNAL_DOWNLOADS]: The skill performs network requests to retrieve data from external websites to extract color palettes and font stacks for brand matching.
Audit Metadata