crm-integrations
Pass
Audited by Gen Agent Trust Hub on Oct 9, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONNO_CODE
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill provides instructions for bidirectional synchronization with external CRM systems (HubSpot, Salesforce, Pipedrive), which creates a potential surface for indirect prompt injection via data ingested from these platforms.
- Ingestion points: CRM records (contacts, leads, deals) and incoming webhook payloads retrieved via API endpoints described in
SKILL.md. - Boundary markers: The instructions do not define specific delimiters or "ignore embedded instructions" warnings to isolate external data from agent instructions.
- Capability inventory: The skill defines actions that modify external state, including
create_deal,update_lead, andconfigure_webhook. - Sanitization: No explicit data validation or sanitization steps are mandated for the agent when processing external CRM content.
- [NO_CODE]: The skill is composed strictly of markdown instructions and YAML metadata. It does not include any accompanying scripts or binary executables, significantly reducing its direct attack surface.
Audit Metadata