decision-maker-outreach
Pass
Audited by Gen Agent Trust Hub on Sep 26, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill instructs the agent to use web search to find relevant industry events and communities. This creates a minor surface for indirect prompt injection, as the agent processes untrusted external data. However, the skill's capabilities are limited to text generation for planning, posing no significant risk of command execution or file system compromise.\n
- Ingestion points: Web search results for communities and events (SKILL.md).\n
- Boundary markers: Not present in instructions.\n
- Capability inventory: Generation of outreach plans and draft messages.\n
- Sanitization: Not specified.\n- [SAFE]: The skill adheres to privacy best practices, explicitly instructing the user and agent to respect anti-spam laws and use only publicly available contact information.
Audit Metadata