reflex-agentkit

The production UI layer for Python AI agents. reflex-agentkit gives an agent a streaming chat, tool-call cards, human-in-the-loop approvals and shared state, in pure Python, on Reflex. It speaks the AG-UI protocol, so it works with PydanticAI, LangGraph and any AG-UI-compatible agent server. There is no TypeScript to write.

from typing import ClassVar

import reflex as rx
from pydantic_ai import Agent

from reflex_agentkit import AgentState, agent_chat
from reflex_agentkit.adapters.pydantic_ai import PydanticAIAgent

travel_agent = Agent("anthropic:claude-sonnet-5-5", instructions="You are a travel assistant.")


@travel_agent.tool_plain(requires_approval=True)
def book_hotel(city: str, nights: int) -> str:
    """Book a hotel room. Requires human approval."""
    return f"Booked {nights} nights in {city}."


class ChatState(AgentState, rx.State):
    agent: ClassVar = PydanticAIAgent(travel_agent)


def index() -> rx.Component:
    return agent_chat(ChatState)


app = rx.App()
app.add_page(index)

What you get

Feature	How
Streaming replies	Markdown that streams token by token, batched to at most one push per 50 ms
Tool calls	A card per call with its arguments, status and result; custom renderers per tool
Approvals	Approve, edit the arguments or reject a tool call before it runs
Questions from the agent	AG-UI standard interrupts, answered with a form you can replace
Shared state	The agent's state mirrored in shared_state, editable by the app
Client tools	Tools the agent calls that run in your Reflex app, for example to draw a chart
Cancellation	A stop button that really stops the agent, within one second
Any AG-UI agent	In process (PydanticAI, LangGraph) or over HTTP (HttpAGUIClient)

Where to go next

• Getting started: install it and run the quickstart.
• Concepts: runs, threads, tool calls, interrupts and shared state.
• Security model: what reaches the browser, and what never does.
• Examples: the quickstart and the five AG-UI Dojo demos for PydanticAI and LangGraph.