goose-video-local
Pass
Audited by Gen Agent Trust Hub on Oct 27, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONPROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill instructions direct the agent to process and execute content from an external 'next_step' field. * Ingestion points: Data from an answer's 'next_step' field (SKILL.md). * Boundary markers: None. The skill lacks delimiters or instructions to ignore embedded malicious content. * Capability inventory: The skill grants the agent permission to execute commands provided by external data. * Sanitization: There is no evidence of sanitization, validation, or escaping of the external input before execution.
- [PROMPT_INJECTION]: The skill uses phrases like 'follow that line exactly' and 'with the command exactly as given' regarding external data. This instruction could be exploited by an attacker to bypass agent constraints if they can control the 'next_step' data source.
Audit Metadata