llm
Pass
Audited by Gen Agent Trust Hub on Mar 21, 2026
Risk Level: SAFEPROMPT_INJECTION
Full Analysis
- [PROMPT_INJECTION]: The implementation pattern provided in the guidelines demonstrates direct interpolation of user-controlled variables into model prompts.
- Ingestion points: The variable
emailAccount.aboutis directly embedded into the prompt string withinSKILL.md. - Boundary markers: While XML-style tags are used, there is no guidance on escaping content to prevent tag breakout.
- Capability inventory: The example code uses a
generateObjectfunction to process these prompts. - Sanitization: No sanitization logic is present in the example code or suggested in the best practices.
Audit Metadata