geo-fix-content
Pass
Audited by Gen Agent Trust Hub on Apr 11, 2026
Risk Level: SAFE
Full Analysis
- [PROMPT_INJECTION]: The skill is designed to process untrusted data from external websites or user input, which presents a surface for indirect prompt injection. Maliciously crafted content could attempt to influence the agent's behavior; however, the skill's highly structured, paragraph-by-paragraph processing instructions provide significant mitigation.
- [EXTERNAL_DOWNLOADS]: The skill retrieves HTML content from external URLs provided by the user to perform its analysis. This is a primary, legitimate function of the tool and is clearly documented.
- [COMMAND_EXECUTION]: There are no suspicious shell commands or unauthorized execution patterns. Installation and usage instructions are consistent with standard development practices and the platform's conventions.
Audit Metadata