README.md

Quantiles Python SDK

Python SDK for the quantiles local AI workload observability server.

Installation

uv add quantiles

Usage

To build a custom eval with Python, use the below code. To ensure this eval is runnable with qt run, set up a quantiles.toml configuration file. See ../CONFIG.md for details.

import asyncio
from quantiles import workflow, step, emit, entrypoint

async def handler(input_value, ctx):
    result = await ctx.step(
        "fetch-data",
        {"url": "https://example.com"},
        lambda: {"status": 200}
    )
    await ctx.emit("latency_ms", 50, "ms")
    return result

my_workflow = workflow("demo", handler)

if __name__ == "__main__":
    entrypoint(my_workflow)

In local development, the SDK executes user code locally. The qt server deduplicates steps, triggers workflows, owns durable state, stored outputs, observability records, and metrics.

Development

Run tests:

mise run test

Run linter:

mise run lint