senior-ml-engineer
Pass
Audited by Gen Agent Trust Hub on Apr 12, 2026
Risk Level: SAFEPROMPT_INJECTION
Full Analysis
- [PROMPT_INJECTION]: The skill documents prompt templates in
references/llm_integration_guide.mdthat interpolate user-provided strings (e.g.,{user_input}) directly into the model's context. This configuration presents a vulnerability surface for indirect prompt injection where malicious content in the processed data could override the intended behavior. \n - Ingestion points: User input and product context are interpolated into prompt templates in
references/llm_integration_guide.md. \n - Boundary markers: The templates do not utilize robust delimiters or explicit instructions to ignore embedded commands within the interpolated data. \n
- Capability inventory: The skill outlines patterns for calling LLM APIs and processing responses, which are susceptible to influence from injected instructions. \n
- Sanitization: No input validation or sanitization logic is implemented for the data before it is formatted into the prompts.
Audit Metadata