Use the current TypeSafe API and confidence guide when implementing requests.
Define the claim and candidate passages; keep expected labels outside model state. Use a Noul for whether a passage fully supports a claim, or a Choice when code must distinguish support, contradiction and silence. Include entity, time, qualifiers and scope. Missing evidence is not a denial. Copy receipts from source spans in code; a verbatim receipt establishes provenance, not correctness. Measure retrieval coverage separately from the judgments on retrieved passages.
Include ordinary cases, missing facts, contradictions, negation, different entities, numerical mismatches, dropped conditions, changes from allowed to required, valid inferences and instructions embedded in documents. Keep synthetic fixtures public and customer inputs private.
Sweep thresholds on development cases and freeze the policy before testing unseen cases. Report accuracy on accepted cases together with review rate, coverage and errors per category. Use balanced accuracy for uneven classes. Neither confidence nor a chosen-label probability is an accuracy guarantee, and repeated calls do not increase the number of independent cases.
Compare against the existing workflow on matching inputs. For an LLM baseline, the System One Adapter accepts the same Python requests. Evaluate escalation to an LLM as a separate arm and count its cost and latency. Pin model versions, alternate the arms, and include retries, malformed outputs and service errors.
Use the user's existing harness. Provide an offline dry run and save answers, returned model versions and usage so policy changes can be replayed without new calls. Deliver a report with case counts, accuracy, coverage, end-to-end latency, cost and representative failures.