affiliate-program-terms
Pass
Audited by Gen Agent Trust Hub on Sep 13, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill is designed with a defensive posture toward legal risks, explicitly requiring disclaimers and markers for human legal review on all generated content. It avoids providing final legal advice and limits itself to drafting assistance.
- [INDIRECT_PROMPT_INJECTION]: The skill processes user-supplied data (such as existing agreements and interview answers) which represents a standard ingestion surface for indirect prompt injection.
- Ingestion points:
SKILL.md(Interview section and revision of current terms documents). - Boundary markers: The skill requires the use of markdown formatting and specific
[LEGAL REVIEW]markers for sensitive generated clauses. - Capability inventory: Capabilities are limited to text generation, research via web browsing, and persistent memory usage. There are no mentions of shell command execution or direct file system access.
- Sanitization: The workflow enforces a mandatory legal review disclaimer and human-in-the-loop oversight before any generated draft is used.
- [EXTERNAL_DOWNLOADS]: The skill includes instructions to browse the web to verify regulatory updates. This behavior is restricted to research purposes and targets reputable regulatory sources, representing normal functional use of an agent's browsing capabilities.
Audit Metadata