1
Ready when needed
Messages and events start active tasks only when an agent has something to do.
Zero-idle agent runtime
Kaenu
Stateful agents without idle runtime.
Personal agents without idle infrastructure. Keep them ready when needed, preserve their state, enforce bounded cost, and release runtime when active tasks finish.
The product wedge
Agents are useful. Running them is still too much work.
Capable agents need webhooks, VMs, state, budgets, secrets, snapshots, monitoring, and cleanup. Kaenu turns that operational burden into a managed runtime for stateful personal agents.
2
Keep portable state
Memory and working context survive between active runtime sessions.
3
Standby stops costing
Runtime is released or destroyed when active tasks finish, with hard budget limits.
The enemy
The future will not lack automation. It will lack aligned automation.
Every platform and institution will offer assistants. Most will still serve the system that owns them. Kaenu is built around agents that act under the user's mandate instead of hidden platform incentives.
Bounded cost
Budgets and idle shutdown are part of the product, not an operations afterthought.
Accountable action
Agent work should leave lifecycle and action evidence the user can inspect.
Honest trust boundaries
The platform must say which providers remain in the trust chain.
Trust model
Built for trust boundaries, not trust theater.
Kaenu isolates tenant runtime, minimizes retained content, keeps state portable, and distinguishes standard runtime from future confidential-compute tiers. It does not pretend Telegram bot transport, external LLM APIs, or standard cloud VMs disappear from the trust chain.
Doctrine
Life Autonomy is the core idea.
Life Autonomy is the practical ability to live according to your own values, responsibilities, relationships, and direction in a machine-mediated world. Kaenu exists to make self-determination possible in that world.