The Agent Skills Directory

[PROMPT_INJECTION] (LOW): Detected an indirect prompt injection surface (Category 8) where untrusted input is processed.
Ingestion points: The emails list and list_name parameters in the upload_users_to_hubspot function (SKILL.md).
Boundary markers: Absent. The skill interpolates these variables directly into JSON payloads for API requests without using delimiters or instruction-isolation markers.
Capability inventory: The skill performs network write operations (POST and PUT requests) to the HubSpot CRM API using urllib.request.
Sanitization: Absent. There is no validation or escaping logic applied to the input strings before they are transmitted to the external service.

hubspot-crm