extract
Pass
Audited by Gen Agent Trust Hub on Sep 15, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill facilitates the ingestion and transformation of external project data which presents an indirect injection surface.
- Ingestion points: Identified in the 'Discover' section of SKILL.md where the agent is instructed to locate and analyze shared UI directories and components using search tools.
- Boundary markers: Absent; there are no specific instructions to treat found code as untrusted or to use delimiters when processing found patterns.
- Capability inventory: The 'Extract & Enrich', 'Migrate', and 'Document' sections grant the skill capabilities to create new components/tokens, replace existing code systematically, and delete old implementations.
- Sanitization: Absent; the skill does not specify validation steps for the code it refactors.
- [SAFE]: The skill focuses on legitimate software engineering tasks (refactoring, documentation, component extraction) without suspicious or hidden behaviors.
- [SAFE]: No evidence of hardcoded credentials, sensitive file exfiltration, or unauthorized command execution was found.
Audit Metadata