REX-03: harness: Rust bench binary driving resident apr serve + deterministic verdict parser + receipt writer + scorer (Wilson, McNemar exact, seeded bootstrap, Holm)

owner: rex
Epic: #4354 (REX-001). Spec: docs/specifications/review-experiment-protocol.md §7 row REX-03.

Done when (all must hold): contract review-experiment-receipt-v1 (pv): missing model sha -> inadmissible; NotRun counted as correct -> rejected; Unparsed counted as PASS -> rejected; scorer unit-tested vs hand-computed fixtures; each falsifier planted RED->GREEN

Binding: §6 R-1..R-12, §8 STOPs. Ticket id = PMAT-.

REX-03 landed locally on rex/001 as 0ab6ce7 (it ships in the single REX PR).

Harness + scorer + review-experiment-receipt-v1: 47/47 lib tests, clippy clean.
FALSIFY-RXR-001/002/003 each planted RED (rc=101) and restored GREEN, source byte-identical.
Real-corpus smoke: 105 NotRun{NoDeclaredExecutor} rows → recall 0/70, rejected 0. A signed score verifies; a tampered one is refused.
Live cells wait on infra#1088 executors.
Receipt: docs/audits/rex-001/rex-03-receipt.md

Folded into #4354 as a checklist line (ISSUE-TREE-001: a family of rows becomes one parent with checklist lines). Work and status are tracked there. aprender-wtix-even

Navigation Menu

Repository navigation

REX-03: harness: Rust bench binary driving resident apr serve + deterministic verdict parser + receipt writer + scorer (Wilson, McNemar exact, seeded bootstrap, Holm) #4358

Metadata

Assignees

Labels

Type

Projects

Milestone

Relationships

Development

REX-03: harness: Rust bench binary driving resident apr serve + deterministic verdict parser + receipt writer + scorer (Wilson, McNemar exact, seeded bootstrap, Holm) #4358

Description

Activity

noahgift commented on Sep 25, 2026

noahgift commented on Oct 4, 2026

Metadata

Metadata

Assignees

Labels

Type

Projects

Milestone

Relationships

Development

Issue actions