1.1.0

First release.

An async Python client for TypeSafe Jev, the System One decision model. Ask typed
questions about a piece of text and get values back rather than prose: a probability
for a yes/no question, one option out of a set with the whole distribution, or a
rating against levels you describe.

• Three question types, mixable in one request.
• instructions and every criteria value accept a string, an object or an array,
  because the model reads JSON field names as labels.
• A choice takes 2 to 255 options and a score takes 2 to 10 levels, both checked
  locally so a mistake costs no request.
• ScoreAnswer.normalized divides by the rubric's own top level, which is what a
  weighted composite needs.
• One exception class per documented status code, and no internal retry loop, so the
  caller owns the policy.

Measured from the Netherlands: three questions took 712 ms and a hundred took 714 ms,
so batching questions into one call is nearly free in time. Question text is billed
as input at roughly 38 tokens each, so it is not free in money.