hubspot
Pass
Audited by Gen Agent Trust Hub on Sep 30, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill references and fetches content from
specs.apideck.com,developers.apideck.com, anddevelopers.hubspot.com. These are verified vendor resources and official service documentation. - [INDIRECT_PROMPT_INJECTION]: The skill facilitates reading data from HubSpot CRM (contacts, deals, activities), which represents an attack surface for instructions embedded in external data.
- Ingestion points: Data retrieved from
apideck.crm.contacts.list,apideck.crm.opportunities.list, and other CRM endpoints inSKILL.md. - Boundary markers: None explicitly defined in the instructions to separate data from commands.
- Capability inventory: The skill allows for full CRUD operations on CRM records and arbitrary API calls via the Proxy API feature.
- Sanitization: There are no instructions for the agent to sanitize or validate the content of retrieved CRM records before processing.
Audit Metadata