Quickstart

Alpha (0.1.x) — MIT licensed. Public API may change before 1.0.

Install

pip install lexigram-workflow

Minimal example — Pipeline

Create a simple multi-step pipeline, wire it via the module, and execute it.

import asyncio

from lexigram import Application
from lexigram.workflow import Pipeline, step, WorkflowModule


async def main() -> None:
    app = Application(name="wf-demo")

    # Wire the workflow subsystem
    app.add_module(WorkflowModule.configure())

    async with Application.boot(name="wf-demo") as app:
        pipeline = Pipeline()

        @step(name="greet")
        async def greet(ctx: dict) -> str:
            return f"Hello, {ctx.get('name', 'world')}!"

        @step(name="shout")
        async def shout(ctx: dict) -> str:
            return str(ctx.get("greet", "")).upper()

        pipeline.add_step(greet)
        pipeline.add_step(shout)

        result = await pipeline.execute({"name": "Lexigram"})
        print(result)  # "HELLO, LEXIGRAM!"


asyncio.run(main())

Minimal example — Graph workflow

For directed-graph workflows with branching and conditional routing:

from lexigram.workflow import WorkflowBuilder, WorkflowEngine
from lexigram.workflow.graph.node import AbstractWorkflowNode


class GreetNode(AbstractWorkflowNode):
    name = "greet"
    async def execute(self, state: dict) -> dict:
        return {"message": f"Hello, {state.get('name', 'world')}!"}


class ShoutNode(AbstractWorkflowNode):
    name = "shout"
    async def execute(self, state: dict) -> dict:
        return {"message": state.get("message", "").upper()}


builder = WorkflowBuilder()
builder.add_node(GreetNode())
builder.add_node(ShoutNode())
builder.add_edge("greet", "shout")

engine = WorkflowEngine(builder.build())
result = await engine.execute("greet", initial_state={"name": "Lexigram"})
if result.is_ok():
    wf_result = result.unwrap()
    print(wf_result.final_state)

Wiring

from lexigram.workflow import WorkflowModule

app.add_module(WorkflowModule.configure())

configure() accepts an optional BulkOperationConfig and optional SagaStoreProtocol for durable saga state. The module exports PipelineProtocol.

Next steps

GUIDE.md — pipelines, graph workflows, sagas, bulk operations
ARCHITECTURE.md — internal design, providers, container bindings
CONFIGURATION.md — every config key with defaults
HOWTOS.md — graph branching, sagas, state persistence, versioning
API.md — full reference