Trust Center
Squillify is a controlled execution layer for AI agents. Here is how we protect your data, secure the platform, and handle microskill execution.
The items below describe default platform behavior. Individual microskills declare their own execution characteristics in their descriptors; review the descriptor of the specific skill you intend to use.
Payloads are not stored
Raw submitted payloads are not permanently stored as part of normal microskill execution.
Metadata-only telemetry
Usage telemetry is designed to record request metadata, not the content of your inputs and outputs.
No training on your payloads
Squillify does not use your microskill payloads to train its own machine learning models.
Tenant isolation
Resources are scoped to the tenant that owns them and cannot be overridden by request input.
Explore our policies
How we protect the platform, isolate tenants, and secure credentials.
What happens to the data you send when you invoke a microskill.
External services and sub-processors involved in operating Squillify.
Terms, privacy, and the agreements that govern your use of Squillify.