personalization-6-buckets
Pass
Audited by Gen Agent Trust Hub on Jul 28, 2026
Risk Level: SAFENO_CODEPROMPT_INJECTION
Full Analysis
- [SAFE]: The skill's instructions are purely informational, providing a strategic framework for sales research without any malicious commands or deceptive content.\n- [NO_CODE]: No scripts, binaries, or configuration files are included with this skill, eliminating risks associated with runtime execution or dependency vulnerabilities.\n- [PROMPT_INJECTION]: The skill documentation identifies a standard attack surface for indirect prompt injection by suggesting the agent process external prospect data (e.g., LinkedIn profiles).\n
- Ingestion points: External URLs for prospect profiles and company research mentioned in SKILL.md.\n
- Boundary markers: None defined within the skill instructions.\n
- Capability inventory: No tools or executable capabilities are associated with the skill; the agent's scope is limited to research and content generation.\n
- Sanitization: Not specified.
Audit Metadata