Clef

Add clef (Clef) as a decision helper/tool for my coding agent.
Keep my coding agent on its current chat model. Clef is not a primary coding/chat model and does not generate chat completions.

Write a small native HTTP helper; there is no SDK or vendor key for this integration:
- POST only to https://api.experientiallabs.ai/v1/systemone.
- Read the Bearer key only from EXPERIENTIAL_API_KEY. If missing, stop and ask me to set it locally; never ask me to paste or log it.
- Do not request any other provider API key or call any other endpoint.
- Send model "clef" exactly, preserving any selected :free spelling.

Implement a native JSON request with model, state (text, object, or array), and questions (a map keyed by my question IDs). Each question needs type and instructions:
- choice: criteria maps options such as accept/review to their descriptions.
- noul: judge a proposition; optional criteria describe true and false.
- score: criteria is an ordered list of level descriptions.
Use a concise diff and test summary as state, not the entire repository. Ask independent questions; an answer is not fed into another question in the same request.
Clef reads at most 65,536 tokens of context, shared by state and all question definitions.
Gateway limits: at most 32 questions per request, 64 options per choice, and 2-10 levels per score. These are platform admission bounds, not provider limits.
Request shape reference (the same SystemOne wire): https://docs.typesafe.ai/introduction/quickstart
Model card: https://huggingface.co/Cloudflare/clef

Read answers by question ID (for example answers.review), and show the decision and reported usage to me. Probabilities and confidence are not guarantees of correctness. Never execute a returned choice, merge a change, or bypass review automatically.
Use a bounded timeout, surface HTTP and transport errors, and do not add automatic retries or Idempotency-Key: an unknown outcome may already be charged. This endpoint does not stream and has no Chat Completions, Responses, or Messages facade.
Prepare the helper and a mocked test first. Ask before making a potentially charged live call; do not claim it worked without a real response. Tell me which files you changed.