thinking-partner
Pass
Audited by Gen Agent Trust Hub on Oct 8, 2026
Risk Level: SAFENO_CODEINDIRECT_PROMPT_INJECTION
Full Analysis
- [SAFE]: The skill package consists exclusively of markdown documentation and instructions for a coaching persona. No executable scripts, binaries, or platform-level configurations were identified that could compromise system security.\n- [NO_CODE]: There are no functional code files, scripts, or dependencies included in the skill. The functionality is delivered entirely through behavioral instructions to the AI model.\n- [INDIRECT_PROMPT_INJECTION]: The skill processes user dialogue and 'workspace artifacts' (notes, journals, decision docs) which constitutes a vulnerability surface for indirect prompt injection. The risk is minimized by the skill's lack of automated execution capabilities.\n
- Ingestion points: User conversation and referenced workspace artifacts (SKILL.md).\n
- Boundary markers: Absent; the skill does not specify delimiters or 'ignore' instructions for external data.\n
- Capability inventory: The skill allows for writing 'small workspace notes' (SKILL.md) but lacks network or system execution tools.\n
- Sanitization: No instructions for sanitizing or escaping user-provided data are included.
Audit Metadata