rebuno

README

A kernel-authoritative execution runtime for AI agents.

Rebuno sits between your agents and the tools they use, giving you policy control, a complete audit trail, and operational visibility over autonomous agent behavior. Agents propose actions. The kernel decides whether they're allowed. Every decision is recorded.

Quick Start

Prerequisites: Go 1.25+, Python 3.10+

Start the kernel:

go install ./cmd/rebuno
rebuno dev

Start an agent:

pip install rebuno
python examples/python/agent/hello.py

Create an execution:

rebuno create --agent hello --input '{"query": "hello world"}'

See the full audit trail with rebuno events {id}.

See Getting Started for the full walkthrough.

Documentation

Doc	Description
Getting Started	Quick start walkthrough
Architecture	Core concepts and state transitions
Deployment	Production setup, auth, and configuration
Python SDK	Building agents and runners in Python
TypeScript SDK	Building agents and runners in TypeScript
Tools	Local, remote, and MCP tools
Policy	Declarative policy rules
API Reference	HTTP endpoints and schemas
Events	Event types and payloads
CLI	CLI reference

License

MIT