lead-routing
Pass
Audited by Gen Agent Trust Hub on Oct 26, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONNO_CODE
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill provides instructions for designing systems that ingest data from untrusted sources, including web forms, API integrations, and list imports. This defines an attack surface where malicious instructions embedded in lead fields could influence the agent's logic. While the skill suggests validation, it lacks explicit prompt-level defense requirements.\n
- Ingestion points: Data from web forms, API, list imports, and chat tools (as defined in the 'Interview' section of SKILL.md).\n
- Boundary markers: The instructions do not prescribe the use of delimiters or explicit warnings for the agent to ignore instructions in the data it processes.\n
- Capability inventory: The designed logic performs CRM owner-field writes and triggers automated notifications/escalations.\n
- Sanitization: General data validation is mentioned, but specific sanitization to mitigate prompt injection is absent from the workflow.
Audit Metadata