Most automation eventually calls a REST API — a cloud provider, a CI system, a webhook. The requests library makes HTTP simple:

• requests.get/post/put/delete(url, ...) for the verbs
• params= for query strings, json= for a JSON body, headers= for auth
• .status_code, .json(), and .raise_for_status() to handle responses

Always check the status and handle failures. raise_for_status() turns a 4xx/5xx into an exception so errors don’t pass silently. Read secrets (tokens) from the environment, never hardcode them.

import os, requests

token = os.environ["API_TOKEN"]            # from the environment
headers = {"Authorization": f"Bearer {token}"}

resp = requests.get(
    "https://api.example.com/v1/servers",
    headers=headers,
    params={"status": "running"},
    timeout=10,
)
resp.raise_for_status()                     # raise on 4xx/5xx
for server in resp.json()["items"]:
    print(server["name"])

# Create a resource
requests.post("https://api.example.com/v1/tags",
              json={"name": "prod"}, headers=headers, timeout=10)

1. GET a public JSON API (e.g. a weather or quotes API) and print one field.
2. Add raise_for_status() and trigger it by hitting a 404 URL.
3. Read an API token from an environment variable instead of hardcoding it.
4. Add a timeout and explain what happens without one.

• ↗ Requests — Quickstart