geo-fix-schema
Pass
Audited by Gen Agent Trust Hub on Apr 11, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill demonstrates robust security practices by including a dedicated 'Security: Untrusted Content Handling' section. This section instructs the agent to treat all fetched content as untrusted data and provides clear guidance on how to avoid following instructions embedded in that data.
- [PROMPT_INJECTION]: Static analysis identified the presence of prompt injection patterns like 'Ignore previous instructions'. However, context shows these are used as examples in defensive instructions to train the agent to recognize and ignore such attacks when processing external web content. This is a security mitigation, not a vulnerability.
- [SAFE]: The skill's operations are limited to network reads for data extraction and writing the results to a structured JSON file. It does not perform any dangerous command execution, use obfuscation, or access sensitive local files.
Audit Metadata