press-release-writing
Pass
Audited by Gen Agent Trust Hub on Apr 14, 2026
Risk Level: SAFEPROMPT_INJECTIONEXTERNAL_DOWNLOADSCOMMAND_EXECUTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: References the installation of the
inference.shCLI tool and additional skill modules from theinference-shGitHub repository. - [COMMAND_EXECUTION]: Instructs the agent to use the
infshcommand-line utility to run research applications such astavily/search-assistantandexa/search. - [PROMPT_INJECTION]: The skill exhibits an indirect prompt injection surface (Category 8) by ingesting untrusted data from external web searches to inform its writing process.
- Ingestion points: External search results retrieved via the
infshCLI tools (SKILL.md). - Boundary markers: Absent; there are no specific delimiters or instructions provided to the agent to ignore or isolate potential instructions embedded in the search results.
- Capability inventory: Execution of the
infshCLI tool for research and fact-checking (SKILL.md). - Sanitization: Absent; the skill does not specify methods for validating or filtering the retrieved data before processing.
Audit Metadata