superdialog 0.2.2

Standalone dialog state machine framework — text in, text out.

SuperDialog

Standalone dialog state machine framework. Pure text in, pure text out.

SuperDialog is the brain layer for conversational systems. It turns a prompt or a flow graph into a running dialog state machine — managing turn-by-turn logic, tool calls, flow transitions, and conversation memory.

User text → DialogMachine.turn() → Agent reply text

Audio, STT, TTS, telephony, and media servers are out of scope — those belong to voice infrastructure like LiveKit, PipeCat, or the Unpod Voice Platform. SuperDialog ends at text in, text out.

SuperDialog is to conversation flow what n8n is to integration workflow — a small, composable, eval-able runtime for orchestrating turn-by-turn logic. Where LangChain and LangGraph expose general agent primitives, SuperDialog focuses narrowly on the conversational state machine: who speaks next, what flow to switch to, when to call a tool, when to escalate.

---

Why standalone

The brain has natural reuse beyond voice. A dialog state machine that runs a customer-onboarding flow works the same whether the user is on a phone, a WhatsApp thread, an Intercom widget, or a CLI test harness. Coupling it to telephony forecloses every non-voice use case.

The dependency direction matters. Voice infrastructure should depend on SuperDialog (as one brain option), not the other way around — keeping the framework portable and the platform composable.

Because the interface is text-only, every dialog is a unit-testable function. No audio fixtures, no API keys, no phone number to test a flow.

Install

pip install superdialog

Install only the extras you need:

pip install superdialog[livekit]    # LiveKit adapter
pip install superdialog[pipecat]    # PipeCat adapter
pip install superdialog[fastapi]    # FastAPI adapter + uvicorn
pip install superdialog[ws]         # WebSocket runner
pip install superdialog[mcp]        # MCP tool support
pip install superdialog[langchain]  # LangChainAgent

Quickstart

import asyncio
from superdialog import create_dialog_flow, DialogMachine, Flow

# 1. Bootstrap a flow from a prompt (one-shot LLM call at construction).
#    The build LLM is used ONLY here — never at runtime.
async def build():
    flow = await create_dialog_flow(
        prompt="Confirm appointment. Ask if Friday 4pm works; offer 5pm if not.",
        llm="openai/gpt-5.1",
    )
    flow.save("appointment.json")        # JSON, version-controllable

asyncio.run(build())

# 2. Build the runtime machine (runtime model can differ from the build model).
dialog_machine = DialogMachine(
    flow=Flow.load("appointment.json"),
    llm="anthropic/claude-haiku-4-5",
)

# 3. Run a conversation.
async def chat():
    reply = await dialog_machine.turn("Hi, I'm calling about my appointment.")
    print(reply.text)

asyncio.run(chat())

Or skip the Python and use the bundled CLI:

superdialog chat appointment.json

Add a tool

Three shapes, one interface — Python callables, HTTP endpoints, MCP servers:

from superdialog import DialogMachine, Flow, PythonTool

def lookup_customer(phone_number: str) -> dict:
    """Look up customer record by phone number."""
    return crm.get_by_phone(phone_number)

dialog_machine = DialogMachine(
    flow=Flow.load("appointment.json"),
    llm="anthropic/claude-haiku-4-5",
    tools=[PythonTool.of(lookup_customer)],   # schema inferred from signature
)

HttpTool(...) and MCPTool(...) plug in the same way. Tool results merge into the flow's slots and can trigger an edge transition.

Model URIs

Pick a provider per machine with a LiveKit/litellm-style URI — and swap it at runtime with dialog_machine.set_llm(uri):

URI	Routes to
openai/gpt-5.1	OpenAI
anthropic/claude-opus-4-7	Anthropic
google/gemini-2.5-pro	Google
groq/llama-3.3-70b	Groq
bedrock/<model>	AWS Bedrock
vllm/<model>@<host>	Self-hosted vLLM
ollama/<model>@<host>	Self-hosted Ollama
openrouter/<vendor>/<model>	OpenRouter
custom/<name>/<model>	Developer-registered via register_llm_provider(...)

Bring your own LLM:

from superdialog import register_llm_provider

register_llm_provider(
    name="internal",
    base_url="https://llm.example.com/v1",
    api_key=os.environ["LLM_KEY"],
    api_style="openai",
)
# usable as "custom/internal/llama-3-70b-tuned"

Deploy anywhere

The same DialogMachine object drops into every host. The host varies; the SuperDialog code is identical.

Host	Adapter	Approx. LoC
CLI	none — superdialog chat or an input()/print() loop	~5
LiveKit	superdialog.adapters.livekit.DialogMachineLLM (Agent(llm=...) plugin)	~6
PipeCat	superdialog.adapters.pipecat.make_processor(dm)	~2
FastAPI	direct /turn route, or a SessionWorker for multi-user	~6
Unpod Voice	superdialog.adapters.websocket.WebSocketRunner	~6
Slack / Discord / IRC / etc.	none — direct callback	~3

# LiveKit — same dialog_machine, ~6 lines
from livekit.agents import Agent, AgentSession
from superdialog.adapters.livekit import DialogMachineLLM

async def entrypoint(ctx):
    agent = Agent(llm=DialogMachineLLM(dialog_machine))
    await AgentSession().start(agent=agent, room=ctx.room)

Sessions

A live DialogMachine holds its conversation in memory for the lifetime of the instance. When a conversation must outlive the process (multi-worker FastAPI, day-long chat), wrap it in a SessionWorker:

from superdialog import DialogMachine, SessionWorker, InMemorySessionStore

worker = SessionWorker(
    agent_factory=lambda: DialogMachine(flow=flow, llm="openai/gpt-5.1"),
    store=InMemorySessionStore(),
)

async with worker.acquire("user-42") as h:
    reply = await h.turn("hello")

The worker multiplexes N concurrent sessions, shares the immutable Flow by reference, and serializes same-session requests via a per-session lock. LLMAgent and LangChainAgent are drop-in non-state-machine brains for the same machinery.

CLI

Command	Purpose
superdialog chat <flow.json>	Interactive terminal chat
superdialog flow lint <flow.json>	Validate graph structure
superdialog flow draw <flow.json>	Render a Mermaid diagram
superdialog flow generate "<prompt>"	Bootstrap a flow.json from a prompt

Feature status

Capability	Status
Prompt → flow, turn execution, model URIs	✅ v0.1
Tools (Python / HTTP / MCP), FlowSet + switch_flow	✅ v0.1
CLI (chat, flow lint / draw / generate)	✅ v0.1
Adapters (LiveKit, PipeCat, FastAPI, WebSocket)	✅ v0.1
SessionWorker — multi-conversation lifecycle + persistence	✅ v0.2
LLMAgent, LangChainAgent — non-DM brains	✅ v0.2
assist(text) — mid-conversation system injection	✅ v0.2
Distributed stores (Redis / File / SQLite) + RedisLockBackend	🔜 v0.3
Eval harness + superdialog eval CLI	🔜 v0.3
True provider-level streaming inference	🔜 v0.4

What it is not

• Not a UI flow designer — that belongs to a downstream tool.
• Not a voice framework — audio, STT, TTS are out of scope.
• Not multi-modal — text only at the interface (vision/audio via tools).
• Not a hosted service — a library. Hosting is offered by the Unpod Voice Platform for those who want it.

Documentation

Doc	Contents
docs/00-overview.md	What it is, why standalone, positioning, roadmap
docs/01-architecture.md	Components, contracts, data shapes
docs/02-api-reference.md	Every class and method
docs/03-embedding-guides.md	Host-by-host integration walkthroughs

License

Apache-2.0. See LICENSE.