lead-enrichment
Pass
Audited by Gen Agent Trust Hub on Oct 10, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted lead lists that are processed by LLMs (Claude or Claygent) for normalization and standardization, creating an indirect prompt injection surface where malicious data could influence agent behavior.\n
- Ingestion points: Phase 1 (Intake) requires users to provide lead lists containing company and person data.\n
- Boundary markers: The skill instructions do not specify the use of delimiters or 'ignore' instructions for the data being processed by the LLM.\n
- Capability inventory: The skill uses LLM capabilities for data normalization, standardized mapping, and gap-filling across multiple providers.\n
- Sanitization: No explicit sanitization or filtering of the input lead data is defined before processing.
Audit Metadata