batna-zopa

Pass

Audited by Gen Agent Trust Hub on Aug 22, 2026

Risk Level: SAFENO_CODE
Full Analysis
  • [SAFE]: The skill consists exclusively of Markdown documentation and instructional prompts. There are no executable scripts (Python, Node.js, etc.) or binary files included.
  • [NO_CODE]: No automated tools or command-line interfaces are invoked by the skill instructions. The 'Coaching Novices' section uses text-based control markers like '[WAIT]' to guide conversation flow, which does not pose a security risk.
  • [SAFE]: External links in the documentation point to the vendor's own infrastructure (deciqai.com), their GitHub organization, or well-known educational domains (pon.harvard.edu). No obfuscated URLs or suspicious remote resources were detected.
  • [SAFE]: The instructions do not attempt to override AI safety guidelines, bypass constraints, or extract system prompts. The content is strictly limited to teaching negotiation techniques.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 22, 2026, 08:14 AM
Security Audit — agent-trust-hub — batna-zopa