Use cases

You're Claude Code, working in my repository.

Task
Pick the use-case recipe on this page (Meta's cookbook, Use cases) that fits this project best, tell me which one and why in two sentences, then build its core loop here from that recipe's own page: the request shape, the tools or harness it uses, and its verification step.

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

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
- The recipes in this section: chart analysis, error screenshot fix, smart glasses with OpenClaw, generating slides, browser-verified web design, iterative game dev, sandboxed execution, multi-agent orchestration, one-shot game dev, perception grounding, GitHub agent, computer use (Cua sandbox), computer use on macOS (metacua), and a local Muse Glimmer agent that asks before calling Muse Spark.
- Every recipe calls https://api.meta.ai/v1 with an OpenAI-compatible client; most use the Responses API (POST /v1/responses).
- Harnesses vary: OpenCode, OpenClaw and Hermes are agent runtimes with a built-in Meta provider; the Python recipes call the API directly. Pick one that runs in this project.

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"])

response = client.responses.create(model="muse-spark-1.3-contributor", input="Say hello in five words.")
print(response.output_text)

Done when
- You named one recipe, linked its page, and said why it fits this project.
- The recipe's main call runs once against the real API and prints what the recipe says it should.
- The recipe's own verification step (a validator, a test, a screenshot check) runs and passes.
- 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.