These guides explain the model behind MyApi rather than the individual endpoints. Start with connected services: how an OAuth connection is made once, how the resulting tokens are encrypted at rest, and how agents borrow that access without ever seeing the underlying credentials.
From there the scope model covers how access narrows from a workspace down to a single service and a single operation, and how per-agent usage limits and resource sub-scopes keep one agent from reaching data that belongs to another. Device approval adds a second gate: a new agent device stays inert until you approve it from the dashboard or your phone.
Later sections cover workspaces and roles for shared setups, automations and triggers for work that should run without being asked, the memory and context layer agents read from, and the audit log that records every call. Each guide ends with the concrete steps to set the feature up and the scopes it requires.