press-release-writing
Pass
Audited by Gen Agent Trust Hub on Sep 8, 2026
Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill uses the
beltCLI tool to execute search and fact-checking queries. Usage is restricted via theallowed-toolsconfiguration to only permitbeltcommands. - [EXTERNAL_DOWNLOADS]: The skill references the installation of the
belt-sh/clitool vianpxand points to documentation hosted on GitHub. These references are used to set up the necessary research infrastructure for the skill's functionality. - [INDIRECT_PROMPT_INJECTION]: The skill ingests external data from search tools (Tavily and Exa) to assist with fact-checking, which creates a potential surface for indirect prompt injection from the retrieved content.
- Ingestion points: Data returned from
belt app runfor search and answer tools. - Boundary markers: None identified in the provided instructions.
- Capability inventory: Shell command execution restricted to the
belttool. - Sanitization: No specific sanitization or filtering of search results is described.
Audit Metadata