HubSpot

**Judgement** A solid, fairly comprehensive HubSpot integration skill that exposes core CRM and key CMS capabilities. It should work well for production-grade assistants that need to read and update HubSpot data, as long as the access token and scopes are managed carefully. **What it does well** - Supports core CRM entities: contacts, companies, deals, owners, and their associations. - Provides search, list, and get-by-ID patterns plus property-schema inspection, enabling dynamic and schema-aware behaviors. - Adds CMS coverage (pages, domains, files), allowing combined CRM+content workflows. - Documentation is concrete and practical (curl + PowerShell, search operators, rate limits, pagination), which should map cleanly to tool calls. **Key risks / limitations** - Uses a powerful private app access token; if that token is over-scoped, an agent could read or modify a wide range of CRM/CMS data. - Notes that full CRUD is supported, implying potential for destructive operations (e.g., updates and possibly deletions) if not constrained by scopes or higher-level policies. - Relies entirely on HubSpot API availability and rate limits (100 requests / 10 seconds); heavy, highly concurrent agents may hit throttling. - Complex HubSpot schemas (many custom properties) can make it harder for agents to choose the right fields without additional schema guidance. **Recommended scenarios** - Sales, CS, or RevOps copilots that need to look up and update contacts, companies, and deals in real time. - Support and account-management agents that need quick 360° context (associated deals, companies, and owners) for a given contact. - Marketing and web ops assistants that must inspect HubSpot CMS pages/domains/files while cross-referencing CRM data. - Internal automation agents where HubSpot private app scopes and guardrails can be tightly controlled by the ops/IT team.