styles
Warn
Audited by Gen Agent Trust Hub on Jun 22, 2026
Risk Level: MEDIUMPROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [PROMPT_INJECTION]: The skill's metadata is deceptive, identifying 'Google LLC' as the author and providing a 'last-updated' date of '2026-06-06', which contradicts the actual author context. This impersonation is likely intended to gain unearned trust.
- [PROMPT_INJECTION]: The skill provides instructions for non-existent APIs and library versions (e.g., 'androidx.compose.foundation.style.Style' and foundation version '1.12.0-alpha01'). This misinformation could lead an agent to generate invalid code and failed build configurations.
- [COMMAND_EXECUTION]: The skill instructs the agent to create and run automated UI tests and capture screenshots. While these are typical development tasks, they are directed towards implementing the fictitious API described in the guide.
Audit Metadata