Cancel a response

You're Claude Code, working in my repository.

Task
Add a Stop button for long requests in my app: start them with background: true, keep the id, cancel with POST /v1/responses/{response_id}/cancel, show the status the cancel returns, and fetch the response with GET when it had already finished.

Read this first
Meta's docs page "Cancel a response" as Markdown: https://dev.meta.ai/docs/api-reference/responses/cancel-response.md
It's the source of truth for request and response shapes. The same page for people: https://dev.meta.ai/docs/api-reference/responses/cancel-response

Setup
- Base URL: https://api.meta.ai/v1
- Auth: send Authorization: Bearer $MODEL_API_KEY, reading MODEL_API_KEY from the environment. Never hard-code, print or commit the key.
- Model: muse-spark-1.3-contributor (Contributor mode: Meta may use what you send to train future models; drop -contributor for Standard mode)
- Python: the OpenAI SDK (pip install openai) with base_url="https://api.meta.ai/v1" and api_key=os.environ["MODEL_API_KEY"] works, or plain requests.

Facts to get right
- Only background responses can be cancelled. The reply has id, object, status, model, background and an always-empty output; its status is the current one (cancelled, or in_progress while the cancel takes hold), so poll GET until it settles; a run can still finish first and settle completed.
- A cancel that races completion returns the completed response; cancelling an already cancelled or failed one returns its current state; a deleted one is a 404.
- The guide says a cancel after completion returns status completed with an empty output[]; the live API has also answered 400 "Cannot cancel a completed response." Handle both, then read the output with GET /v1/responses/{id}.
- background: true can't be combined with stream: true; a background body is capped at 1 MiB.

Starter code (Python)
import os

import openai
from openai import OpenAI

client = OpenAI(base_url="https://api.meta.ai/v1", api_key=os.environ["MODEL_API_KEY"])

job = client.responses.create(model="muse-spark-1.3-contributor", input="Write a 600-word story.", background=True)
try:
    result = client.responses.cancel(job.id)
    print(result.status)  # "cancelled", or "completed" if it won the race
except openai.BadRequestError as err:
    print("too late:", err.message)
print(client.responses.retrieve(job.id).status)

Done when
- Stopping a running background response shows status cancelled; stopping one that already finished shows its output from GET.
- The key comes from MODEL_API_KEY and never appears in code, logs or commits.
- API errors (4xx, 429) surface their message instead of failing silently.