list-building
Pass
Audited by Gen Agent Trust Hub on Sep 28, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [COMMAND_EXECUTION]: The skill instructions include the use of shell commands like
lsto interact with the local file system to retrieve existing campaign data. - [INDIRECT_PROMPT_INJECTION]: The skill processes data from multiple external and user-supplied sources, which introduces a potential for instructions embedded in data to influence agent behavior.
- Ingestion points: The skill ingests data from HubSpot CRM, Clay enrichment services, and user-provided CSV files (documented in SKILL.md).
- Boundary markers: No specific delimiters or "ignore embedded instructions" warnings are defined for the data ingestion steps.
- Capability inventory: The agent has the capability to execute shell commands (
ls) and write JSON files to the localcampaigns/directory. - Sanitization: The instructions do not describe any sanitization, validation, or escaping of the external data fields before they are used to populate account JSONs or influence the targeting workflow.
Audit Metadata