core/evals/constraint_oracle
Shay a098519bc1 feat(constraint): R2 setup oracle + gold + ADR-0211 — the ruler (R2 C2)
evals/constraint_oracle: span-free canonical setup signature (unknowns/facts/constraints/query; terms merged+sorted, lhs constant folded into rhs), a 13-fixture gold (7 solved / 3 solver_refuses / 3 reader_refuses) with a closed expect+refusal taxonomy, and a gold-validation runner. No reader yet — this proves the ruler is coherent (13/13 valid), not capability.

ADR-0211 ratifies the R2 finite-integer two-category constraint compiler: the IR, the canonical signature, the closed taxonomy, the four wrong=0 gates (setup oracle / reader refusal / exact solver / answer-choice contradiction flag), off-serving disjointness, and the C1-C9 ladder. Cross-refs ADR-0207/0175/0083/0055-0057; reuses the R1 setup_oracle pattern. Off-serving (no generate.derivation / reliability_gate). 13 oracle tests incl. per-branch meaningful-fail.
2026-06-07 07:19:29 -07:00
..
__init__.py feat(constraint): R2 setup oracle + gold + ADR-0211 — the ruler (R2 C2) 2026-06-07 07:19:29 -07:00
__main__.py feat(constraint): R2 setup oracle + gold + ADR-0211 — the ruler (R2 C2) 2026-06-07 07:19:29 -07:00
r2_gold.jsonl feat(constraint): R2 setup oracle + gold + ADR-0211 — the ruler (R2 C2) 2026-06-07 07:19:29 -07:00
runner.py feat(constraint): R2 setup oracle + gold + ADR-0211 — the ruler (R2 C2) 2026-06-07 07:19:29 -07:00
signature.py feat(constraint): R2 setup oracle + gold + ADR-0211 — the ruler (R2 C2) 2026-06-07 07:19:29 -07:00