interface-design
Pass
Audited by Gen Agent Trust Hub on Sep 27, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill implements a project-specific memory system by reading and writing design patterns to a local file, creating a surface where untrusted input could influence future agent actions.
- Ingestion points: The agent ingests user design prompts and existing project files (including SKILL.md and system.md) to extract and store design tokens, signatures, and patterns.
- Boundary markers: The instructions do not define explicit delimiters or instructions to ignore embedded commands within the ingested design requirements or code comments.
- Capability inventory: The skill is designed to read from and write to the local filesystem (specifically
.interface-design/system.md) to maintain design state between sessions. - Sanitization: There are no documented procedures for sanitizing or validating the extracted data before it is persisted to the project's memory file, allowing potential injection into the design system documentation.
Audit Metadata