gemini-api-dev
Pass
Audited by Gen Agent Trust Hub on Apr 16, 2026
Risk Level: SAFEPROMPT_INJECTION
Full Analysis
- [PROMPT_INJECTION]: The skill contains directives aimed at overriding the AI agent's built-in training and knowledge. Evidence includes phrases like 'These rules override your training data' and 'Your knowledge is outdated'.
- [PROMPT_INJECTION]: The instructions command the agent to treat established model versions as legacy in favor of unreleased versions, which may cause functional issues or lead to development errors based on inaccurate information.
Audit Metadata