hundred-million-offers
Pass
Audited by Gen Agent Trust Hub on Aug 23, 2026
Risk Level: SAFENO_CODE
Full Analysis
- [PROMPT_INJECTION]: No malicious instructions, bypass markers, or DAN-style injections were detected. The skill provides structured guidance for offer analysis and positioning.
- [DATA_EXPOSURE_AND_EXFILTRATION]: No access to sensitive files, hardcoded credentials, or network exfiltration patterns found. The documentation includes standard Amazon affiliate links for literature referenced in the text.
- [REMOTE_CODE_EXECUTION]: There are no scripts, package installers, or command-line execution patterns present in the skill files. All content is markdown-based.
- [INDIRECT_PROMPT_INJECTION]: The skill processes user-supplied business offers. Ingestion points: User mentions of offer details in SKILL.md. Boundary markers: Absent. Capability inventory: None (no shell access, network, or file writing). Sanitization: Absent. While this creates an attack surface for data ingestion, the skill lacks the capabilities required to facilitate a malicious exploit chain.
- [SAFE]: The skill consists exclusively of markdown documentation, business rubrics, and diagnostic tables for entrepreneurship training.
Audit Metadata