evidence-based-citations
Pass
Audited by Gen Agent Trust Hub on Jun 19, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill is comprised of standard Markdown instructions and a manifest file. No malicious code, obfuscation, or unauthorized access patterns were detected.
- [PROMPT_INJECTION]: The skill instructions provide a robust framework for handling external data. By requiring a strict output format (Field, Value, Quote, Source) and verbatim quotes, it inherently mitigates the risk of indirect prompt injection from fetched content being interpreted as instructions by the agent.
- [COMMAND_EXECUTION]: The skill encourages the use of an agent's built-in browser or fetch tools to retrieve documentation. This is a legitimate use of the tool aimed at improving the accuracy and reliability of the agent's responses.
- [DATA_EXFILTRATION]: While the skill involves network operations (fetching URLs), these are directed towards primary sources like official documentation and RFCs to fulfill user requests for evidence. There is no evidence of sensitive data being transmitted to unauthorized third parties.
Audit Metadata