Motivational Interviewing
Pass
Audited by Gen Agent Trust Hub on Oct 27, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: No malicious patterns or security vulnerabilities were detected across the analyzed files.
- [SAFE]: The skill's functionality is limited to text analysis and response generation based on the principles of Motivational Interviewing. It uses a restricted set of tools (Read, Write, WebSearch) and does not contain any executable scripts or complex automation.
- [SAFE]: External references point to legitimate professional organization websites (motivationalinterviewing.org) and internal markdown guides. No suspicious downloads or remote code execution patterns were found.
- [INDIRECT_PROMPT_INJECTION]: The skill processes user-supplied conversation excerpts through its sub-commands (e.g.,
/mi,/mi:oars). While this presents a theoretical attack surface, the risk is negligible due to the lack of high-privilege capabilities and the purely linguistic nature of the task. - Ingestion points: User statements and conversation logs provided for analysis in
SKILL.md. - Boundary markers: Not present for the input data, but the agent's output is structured by defined templates.
- Capability inventory: Limited to standard file operations and web searching; no dynamic code execution or direct network access available.
- Sanitization: Not present, though unnecessary for the current capability scope.
Audit Metadata