site
Pass
Audited by Gen Agent Trust Hub on Sep 23, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill instructs the agent to build websites using JSON payloads that include user-controllable strings for page content. There is no mention of sanitizing these inputs or using delimiters to prevent injection. * Ingestion points: JSON args in create-site action. * Boundary markers: Absent. * Capability inventory: Uses arc afs exec for file and service management. * Sanitization: Absent.
- [COMMAND_EXECUTION]: The skill directs the agent to execute arc CLI commands to manage the website lifecycle, diagnostics, and deployment within the AFS environment.
Audit Metadata