connect-hubspot-mcp

Pass

Audited by Gen Agent Trust Hub on Jul 13, 2026

Risk Level: SAFEPROMPT_INJECTION
Full Analysis
  • [SAFE]: The skill documents the setup of HubSpot's official remote MCP server at mcp.hubspot.com. All referenced domains and CLI tools (such as hs mcp setup) are associated with HubSpot, a well-known service provider, and the configuration steps follow standard platform procedures.
  • [PROMPT_INJECTION]: The skill facilitates the processing of external CRM data, which constitutes an indirect prompt injection surface. This is an inherent risk of the skill's primary purpose of allowing natural-language interaction with a CRM.
  • Ingestion points: CRM objects including contacts, deals, and companies fetched from mcp.hubspot.com as described in Stage 2 and 3 of SKILL.md.
  • Boundary markers: Not provided; the instructions do not include specific delimiters or 'ignore' instructions for data processed from the CRM to prevent the agent from following instructions embedded in CRM records.
  • Capability inventory: The skill enables read/write access to CRM objects, analytics data, and content creation (landing pages) via MCP tools.
  • Sanitization: The skill does not mention specific sanitization or validation logic for the external data, relying on underlying platform safety controls.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 13, 2026, 12:28 AM
Security Audit — agent-trust-hub — connect-hubspot-mcp