grove-ui-design
Pass
Audited by Gen Agent Trust Hub on Oct 9, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADSCOMMAND_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill implements components such as
GroveTextandGroveTermdesigned to parse and render dynamic content using a special[[term]]syntax. If the input data (e.g., FAQ items, feature lists, or pricing descriptions) originates from untrusted sources, it could serve as a vector for indirect prompt injection. The skill lacks explicit boundary markers or sanitization instructions for this processed content. - Ingestion points:
contentprop inGroveTextandtermprop inGroveTermcomponents withinSKILL.md. - Boundary markers: None present in the provided instructions.
- Capability inventory: Svelte UI rendering and terminology swapping.
- Sanitization: None observed for the interpolated strings.
- [EXTERNAL_DOWNLOADS]: The documentation suggests fetching icon path definitions from the official Lucide repository on GitHub (
raw.githubusercontent.com/lucide-icons/lucide/main/icons/). This is a well-known and trusted source for UI assets. - [COMMAND_EXECUTION]: The skill provides instructions for using a local tool named
glimpseviauv runfor component auditing and visual verification (e.g.,uv run --project tools/glimpse glimpse showroom). These commands are used for design compliance and visual testing within the local development environment.
Audit Metadata