I'm writing a prompt and have no application yet

There is no application here: a prompt in prompts/, five questions in cases.json, and the answers they ought to get. That is enough to tell whether an edit made things better or only different — the question you have on day one, and the one the diff cannot answer.

Two checks, because one is not enough. Levenshtein says how far the answer is from the one you would have written; LlmRubric says the part no string comparison reaches, wrapped in Repeated because a judge asked twice does not answer twice the same.

report.html is what happened when one line — "Always remind the customer of the returns policy" — was added to prompts/system.txt. The report shows the diff of the prompt itself, above the scores it moved: five answers longer, ten checks worse. The prompt is recorded in every run, so the baseline carries the prompt that produced it.

$ uv sync && uv run digline run --suite suite.py
$ uv run digline promote --suite suite.py --run latest
# edit prompts/system.txt, then
$ uv run digline compare --suite suite.py --run latest

No API key: the provider and the judge are stand-ins. DIGLINE_LIVE=1 uses the real ones.