define-design-guidelines
Pass
Audited by Gen Agent Trust Hub on Oct 15, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes external, user-provided files (
CUSTOMER.mdandBRAND_GUIDELINES.md) and interpolates their content into prompts for subagents and user interaction flows. - Ingestion points: Content is read from
CUSTOMER.mdin Phase 0 and Phase 4, andBRAND_GUIDELINES.mdin Phase 4, to inform design recommendations and audit the generated output. - Boundary markers: The instructions lack explicit delimiters or 'ignore' instructions to prevent the agent from obeying malicious commands potentially embedded within the customer or brand documentation.
- Capability inventory: The skill has the ability to read files, launch specialized subagents (
design-researchanddesign-quality-auditor), interact with the user, and write files to the local directory. - Sanitization: No logic is provided to sanitize, escape, or validate the content of the ingested files before it is used to influence agent behavior or document generation.
Audit Metadata