Architecture
How the pieces of Imaginne fit together — surfaces, Engine, Gateway, and Platform — and why the agent runs on your machine.
Local execution & privacy
What runs on your machine, what leaves for the model, and how Imaginne reduces the leakage of secrets and personal data.
Sessions
What a session is, how the turn cycle works, how context is managed, and how to resume conversations.
Memory
Imaginne's two memories — the four per-project notes on the local surfaces and the account memory in the web chat — how the agent uses them and how they are protected.
Models
Imaginne's models — nnumbers and nnumbers-code — automatic selection, and how your organization controls what shows up in the picker.
Autonomy & permissions
The autonomy levels, the command execution mode, and the vault that protects your workspace.
Skills
What a skill is, where skills come from, how the agent triggers them, and what "protected skill" means.
Identity & login
How the single organization login works, the authenticated session that works across every interface, and why you don't manage API keys.
Organizations & policies
The organization as a tenant, and how policy governs models, credentials, skills, and the agent's behavior.