queue-design
Pass
Audited by Gen Agent Trust Hub on Jul 9, 2026
Risk Level: SAFEPROMPT_INJECTION
Full Analysis
- [PROMPT_INJECTION]: The skill workflow involves ingesting data from external web sources, which serves as a potential surface for indirect prompt injection.
- Ingestion points: The
WebFetchandWebSearchtools are utilized inSKILL.mdto gather technical context from external sources. - Boundary markers: The instructions do not define explicit delimiters or warnings to ignore instructions that might be embedded within fetched content.
- Capability inventory: The skill enables access to several tools, including
Bash,Write,Read,Glob, andGrep. - Sanitization: There is no evidence of sanitization or input validation for the data retrieved from external web sources.
Audit Metadata