thinking-method-selector
Pass
Audited by Gen Agent Trust Hub on Oct 6, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill is purely instructional, providing guidance to the agent on how to categorize user problems and apply decision-making frameworks. It does not include any scripts, external dependencies, or network-enabled commands.
- [INDIRECT_PROMPT_INJECTION]: The skill instructs the agent to read repository data or external documentation to frame user problems. While this represents a standard processing surface for data that could contain hidden instructions, the skill lacks sensitive capabilities (such as file writes, credential access, or network exfiltration) that would be required to exploit this surface maliciously.
Audit Metadata