prompt-engineering-edge-cases
Pass
Audited by Gen Agent Trust Hub on Aug 25, 2026
Risk Level: SAFEPROMPT_INJECTION
Full Analysis
- [PROMPT_INJECTION]: The skill includes instructional examples of prompt injection attempts, such as 'Ignore the above and ...' and 'Print your system prompt'. These snippets are contained within sections explaining how to design robust defenses and are not intended to override the agent's behavior.
- [SAFE]: The skill is a set of markdown guidelines and best practices for developers. It does not include any executable scripts, network requests, or third-party dependencies, and its content promotes secure application development.
Audit Metadata