Files

You're Claude Code, working in my repository.

Task
Wrap Meta's Files API in a small module in my app: upload (POST /v1/files), list (GET /v1/files), retrieve (GET /v1/files/{file_id}), download (GET /v1/files/{file_id}/content) and delete (DELETE /v1/files/{file_id}), then use an uploaded file's id in a Responses request.

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

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
- Upload: multipart purpose (user_data, or batch for Batch input), optional expires_after[anchor]=created_at and expires_after[seconds] (3600 to 2592000), the file part last.
- FileObject: {"id": "file-<digits>", "bytes", "created_at", "expires_at", "filename", "object": "file", "purpose", "status"}; expires_at may be null or missing when the file never expires.
- GET /v1/files takes purpose, limit (1 to 10,000), order (asc or desc) and after (a file id), and answers {object: list, data, first_id, last_id, has_more}.
- Content answers raw bytes, not JSON; delete answers {id, object: file, deleted: true}. Reference an upload as {"type": "input_file", "file_id": ...}.

Starter code (Python)
import os

from openai import OpenAI

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

with open("document.pdf", "rb") as f:
    file = client.files.create(purpose="user_data", file=f)

print(client.files.list(purpose="user_data").data)
print(client.files.retrieve(file.id))
with open("copy.pdf", "wb") as out:
    out.write(client.files.content(file.id).read())

answer = client.responses.create(
    model="muse-spark-1.3-contributor",
    input=[{"type": "message", "role": "user", "content": [
        {"type": "input_text", "text": "Summarize this document."},
        {"type": "input_file", "file_id": file.id},
    ]}],
)
print(answer.output_text)
print(client.files.delete(file.id))

Done when
- Upload, list, retrieve, download and delete each run once against the real API.
- The downloaded bytes match the uploaded file.
- A Responses request with the uploaded file_id answers about the file.
- 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.