py-oidc-auth-client 2602.0.2

Client library for the OIDC Auth server.

Typed client library for the authentication routes exposed by py-oidc-auth.

py-oidc-auth-client is the counterpart of the server-side library py-oidc-auth.

While py-oidc-auth helps you add OpenID Connect login, token, and device endpoints to web frameworks, py-oidc-auth-client consumes those routes and gives you ready-to-use bearer tokens for calling protected APIs.

Features

• One high level helper: authenticate()
• Device flow for headless sessions
• Authorization code flow for interactive logins
• Token caching and refresh token support via a token file
• Fully typed public API

Install

python -m pip install py-oidc-auth-client

Import name is py_oidc_auth_client:

from py_oidc_auth_client import authenticate

Relationship to py-oidc-auth

A typical py-oidc-auth server exposes endpoints similar to:

• GET /auth/v2/login
• GET /auth/v2/callback
• POST /auth/v2/token
• POST /auth/v2/device
• GET /auth/v2/logout
• GET /auth/v2/userinfo

This client calls the relevant routes (token and device, and possibly login/callback) and returns a Token object that contains a ready-made Authorization header.

Quick start

from py_oidc_auth_client import authenticate

token = authenticate(host="https://auth.example.org")

# Use with any HTTP client
headers = token["headers"]
print(headers["Authorization"])

Use with httpx

import httpx
from py_oidc_auth_client import authenticate

token = authenticate(host="https://auth.example.org")

with httpx.Client() as client:
    r = client.get("https://service.example.org/protected", headers=token["headers"])
    r.raise_for_status()
    print(r.json())

Token persistence

By default, the client stores tokens in a cache file so you do not have to re-authenticate on every run. You can control where tokens are stored with token_file:

from py_oidc_auth_client import authenticate, TokenStore

token = authenticate(
    host="https://auth.example.org",
    store=TokenStore(path="~/.cache/py-oidc-auth-client/token.json"),
)

You can also point to a token file via environment variable:

• OIDC_TOKEN_FILE

Interactive and non-interactive environments

The client tries to select a suitable strategy:

1. Use a valid cached access token.
2. Refresh using the refresh token.
3. If interactive authentication is possible, fall back to an interactive login.
4. If running in a non-interactive session without a usable token, raise an error telling you how to provide a token file.

For headless sessions, the device flow is the recommended approach.

Advanced usage

If you need more control than authenticate(), use the flow helpers from py_oidc_auth_client.auth.

Device flow

import asyncio
from py_oidc_auth_client import Config, DeviceFlow

async def main() -> None:
    cfg = Config(host="https://auth.example.org")
    flow = DeviceFlow(config=cfg, token=None, timeout=600)

    device = await flow.get_device_code()
    print("Open:", device.uri)
    print("Code:", device.user_code)

    await flow.poll(device["device_code"], int(device["interval"]))
    print(flow.token["headers"])

asyncio.run(main())

Authorization code flow

import asyncio
from py_oidc_auth_client import Config, CodeFlow

async def main() -> None:
    cfg = Config(host="https://auth.example.org")
    flow = CodeFlow(config=cfg, token=None, timeout=120)
    await flow.login()
    print(flow.token["headers"])

asyncio.run(main())

Contributing

Contributions are welcome. Please open an issue to discuss larger changes before submitting a pull request.