rename package doc_agent -> docfield, distribution -> docfield-extract

eval: tick F12, correct FC-1's scope, document FC-3

eval: add --retry-errors to re-predict only the entries that failed

validation: monetary tolerance scales with rounded terms, not with value

eval: correct FC-1's corpus claim, log FC-2, note cent-exact severity

eval: document failure case FC-1 and log its open questions

eval: report documents that never reached the model, grouped by cause

eval: add tuning/held-out split selection to the score command

eval: score under current rules with --revalidate, and detect cache drift

docs: rename to DocField Extract, reframe as key-value extraction pipeline

docs: correct stale money-comparison docstring in eval.normalize

phase 5: evaluation harness (SROIE)