hubspot

Pass

Audited by Gen Agent Trust Hub on Apr 8, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: No security issues detected.
  • [DATA_EXFILTRATION]: The skill communicates exclusively with official HubSpot API domains (api.hubapi.com) to manage CRM objects. It follows best practices by using environment variables (HUBSPOT_ACCESS_TOKEN) for authentication rather than hardcoding credentials.
  • [COMMAND_EXECUTION]: The skill provides examples using standard, well-known tools like curl, jq, and PowerShell's Invoke-RestMethod. It does not download or execute untrusted remote scripts.
  • [INDIRECT_PROMPT_INJECTION]: While the skill ingests data from external CRM records which could theoretically contain malicious instructions, the skill's primary function is data retrieval and management using structured API calls. There are no high-risk execution sinks (like eval or exec) that would allow such data to influence the underlying system.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 8, 2026, 02:44 PM
Security Audit — agent-trust-hub — hubspot