crm-automation

Pass

Audited by Gen Agent Trust Hub on Sep 14, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process data from various untrusted external sources, creating a surface for indirect prompt injection.
  • Ingestion points: External data enters the system through webhooks (form submissions), CRM queries (HubSpot, Salesforce, Pipedrive), and enrichment APIs (Clearbit, LinkedIn).
  • Boundary markers: The skill lacks explicit delimiters or instructions to ignore embedded commands within the interpolated data (e.g., Lead Data: {lead_data}).
  • Capability inventory: The skill has broad permissions, including writing to multiple CRMs, sending automated emails to leads, and posting messages to Slack channels.
  • Sanitization: There is no evidence of data sanitization, validation, or escaping logic applied to external inputs before they are used in AI prompts or outreach templates.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 14, 2026, 04:48 PM
Security Audit — agent-trust-hub — crm-automation