list-clearance
Pass
Audited by Gen Agent Trust Hub on Sep 30, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted lead list data from CSV files and Clay tables (Step 3) and incorporates this content into markdown reports and CSV outputs (Step 7). This creates a vulnerability surface where malicious content within the lead list could attempt to influence the agent or downstream systems.
- Ingestion points: Lead list data is read from local CSV files or remote Clay tables (SKILL.md).
- Boundary markers: The instructions lack specific delimiters or instructions to neutralize or ignore embedded instructions within the ingested data.
- Capability inventory: The agent has permissions to read/write files and execute the vendor-specific
clayCLI tool. - Sanitization: The skill performs basic normalization on email and domain fields but lacks general sanitization for the remaining record content before it is interpolated into generated report files.
Audit Metadata