phone-verification-waterfall

Pass

Audited by Gen Agent Trust Hub on Sep 17, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted data from HubSpot contact fields (names, job titles, and company names) and passes them to external enrichment tools. This constitutes an indirect prompt injection surface as malicious input within the CRM could potentially influence the agent's execution or data processing.\n- Ingestion points: Lead data is fetched from HubSpot via the hubspot_search_contacts tool as defined in Stage 1 of the workflow in SKILL.md.\n- Boundary markers: There are no boundary markers or delimiters specified to isolate contact-sourced strings from the rest of the prompt or tool parameters.\n- Capability inventory: The skill possesses the ability to read from and update HubSpot records (hubspot_update_contact), and communicate with Apollo and Clay APIs.\n- Sanitization: The skill lacks explicit sanitization or validation steps for the data retrieved from HubSpot before it is used in downstream API calls.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 17, 2026, 09:57 AM
Security Audit — agent-trust-hub — phone-verification-waterfall