agent-dashboards
Pass
Audited by Gen Agent Trust Hub on Nov 9, 2026
Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill instructs the agent to use the 'leverage' CLI tool for checking authentication status and publishing artifacts to a remote space.
- [EXTERNAL_DOWNLOADS]: The documentation describes using 'npx' to install additional skill dependencies from the vendor's repository.
- [INDIRECT_PROMPT_INJECTION]: The skill design involves the agent reading and acting upon comments left by teammates on published dashboards, which provides an ingestion point for untrusted instructions. 1. Ingestion points: User comments on published HTML artifacts as mentioned in SKILL.md. 2. Boundary markers: None specified in the instructions for delineating or ignoring embedded commands in comments. 3. Capability inventory: Subprocess execution of the 'leverage' CLI tool and 'npx' commands. 4. Sanitization: No explicit content filtering or sanitization steps are defined for the ingested commentary.
Audit Metadata