RouteKitAI¶
A minimal, production-ready framework for building and orchestrating AI agents.
RouteKitAI provides graph-based orchestration, built-in tracing, and deterministic replay so you can build, observe, and test agents with confidence.
Features¶
- Graph-native orchestration — Workflows as explicit graphs with clear control flow
- Automatic tracing — Every run produces an immutable trace log
- Deterministic replay — Reproduce any run for testing and debugging
- Policy system — ReAct, Supervisor, Graph, and custom policies
- CLI tools — Run, trace, replay, and analyze from the command line
Quick start¶
pip install RouteKitAI
import asyncio
from routekitai import Agent
from routekitai.providers.local import FakeModel
from routekitai.core.tools import EchoTool
model = FakeModel(name="test")
model.add_response("Hello from RouteKitAI!")
agent = Agent(name="my_agent", model=model, tools=[EchoTool()])
async def main():
result = await agent.run("Hello!")
print(result.output.content)
asyncio.run(main())
Documentation¶
- Getting started — Installation, first agent, first graph
- Guides
- Running agents —
run,run_stream, Runtime, policies - Creating tools — Define tools with Pydantic, permissions, redaction
- Graph workflows — Graphs, nodes, edges, GraphPolicy
- Tracing and replay — Trace IDs, CLI, replay for tests
- CLI reference — All
routekitaicommands - Providers — FakeModel, OpenAI, env vars
- Architecture — Design, execution model, and core components
- Security & Governance — Policy hooks, PII redaction, tool filtering, and best practices
- API reference — Main classes and modules