OS for Agent

In Kleinia, agents are first-class citizens of the system, just like processes. Your favourite agents and traditional apps get system-level identity, quotas, permissions and context without any modification.

To a conventional operating system an agent is just a process: keys, memory and permissions are left to each application, so nothing is shared and nothing is managed by the system. In Kleinia the system provides these things directly, as it always has for processes.

Zero-modification onboarding

No changes to your apps: Kleinia recognizes and manages them for you, and they gain system-level scheduling, memory, quotas and permission management. Start opencode, Claude Code and the rest as usual; they show up in agenttop. See Run your agents.

Unified model gateway

One endpoint for every model call: no juggling keys and credit limits, with model permissions and quotas enforced by the system. The session environment points model endpoints at the gateway, which speaks the OpenAI and Anthropic APIs; local and cloud models both plug in. Tokens are metered per agent, so no agent quietly uses up your credit. See Connect models.

One address space

Everything is an agent, and every agent has its own agent:// address, so apps, system services, even agents on another device can be found and called directly. See anything, anywhere, all as agent.

Context and memory

Context is persisted by the system, so agents carry project conventions and past decisions across sessions. Memory grows like a mind: based on PISA, the system does not just store knowledge but updates, evolves and creates knowledge schemas. Memory is stored and managed by the system, and can be shared with other agents as far as you allow.

Observability

Every task leaves a full execution trace: model calls, tool calls and system actions, step by step, with their time and token use. agenttop shows each agent’s resource use live. See Monitor agents.

More system-level abilities

• Tiered permissions. Low-risk actions run automatically, medium-risk ones notify you, and deleting, sending data out or changing the system needs your confirmation. Everything is audited. See Permissions and audit.
• Permissions per task. An agent gets only what its task needs, and loses it when the task ends.
• Suspend and resume. Long tasks can be suspended to free resources and picked up where they left off.
• Scheduled per agent. Foreground agents get low latency and background work gets throughput: resources follow agents, not processes.

Next

Proposed in ACOS and arriving in later releases: agent:// reaching your other devices so tasks run where they run best, heavy work moving to the strongest machine, and interfaces that assemble around what an agent needs.