referral-program
Pass
Audited by Gen Agent Trust Hub on Apr 11, 2026
Risk Level: SAFENO_CODE
Full Analysis
- [SAFE]: The skill consists entirely of instructional markdown files providing frameworks, metrics, and case studies for marketing. No security threats or malicious patterns were identified in the text or metadata.\n- [NO_CODE]: No executable scripts, binaries, or platform-specific command execution patterns were found. The skill relies on natural language instructions for the AI to assist the user.\n- [DATA_EXPOSURE_AND_EXFILTRATION]: No sensitive data access, credential harvesting, or unauthorized network communication patterns were detected. Mentions of third-party platforms like Stripe or Rewardful are purely informational.\n- [INDIRECT_PROMPT_INJECTION]:\n
- Ingestion points: The skill instructions suggest reading a local file
.claude/product-marketing-context.mdto inform its advice (SKILL.md).\n - Boundary markers: No specific delimiters or warnings to ignore embedded instructions are present.\n
- Capability inventory: The skill files contain no subprocess calls, file-write operations, or network requests.\n
- Sanitization: No input validation or sanitization is performed on the ingested context.\n
- Risk Assessment: While an injection surface exists in the ingested context file, the total lack of actionable capabilities or tools prevents any potential malicious instructions from being executed.
Audit Metadata