news-search
Pass
Audited by Gen Agent Trust Hub on Jul 2, 2026
Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill utilizes the
newsjackcommand-line interface to perform news queries and check authentication status (e.g.,newsjack auth status,newsjack news search). This is the intended operational mode for the skill.\n- [PROMPT_INJECTION]: The skill processes untrusted third-party content from the web, creating an indirect prompt injection surface.\n - Ingestion points: News article headlines, metadata, and page content retrieved via the
newsjackCLI or browser search (SKILL.md).\n - Boundary markers: The instructions do not define explicit delimiters or 'ignore' instructions for the retrieved content.\n
- Capability inventory: The agent has the ability to execute shell commands via the
newsjackCLI and perform web searches (SKILL.md).\n - Sanitization: There is no mention of sanitizing or validating the article content before it is presented to the user or other skills.
Audit Metadata