brand-intake
Pass
Audited by Gen Agent Trust Hub on Sep 29, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONNO_CODE
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill identifies an attack surface by ingesting untrusted data from external sources to generate instruction files.
- Ingestion points: Files in the
examples/folder and user-provided website links are analyzed to populateMOTION.md. - Boundary markers: There are no explicit instructions to the agent to ignore potential commands embedded within the reference images or website content.
- Capability inventory: The skill's capabilities are restricted to writing markdown files (
brand.md,MOTION.md,CLAUDE.md) and do not include code execution, network exfiltration, or privileged system access. - Sanitization: No sanitization or filtering of the content derived from external sources is performed before it is written to the project files.
- [NO_CODE]: The skill is composed entirely of natural language instructions in markdown format. It does not contain any executable scripts (e.g., Python, JavaScript, Bash) or binary dependencies, which significantly limits the potential for traditional technical exploits.
Audit Metadata