[cschanhniem]

Implements Q2 from NEXT_100_PR_MAP.

What

• src/openamp_foundry/evidence/wetlab_hit_record.py: WHR- schema; WetlabHitRecord dataclass (16 fields); validate_wetlab_hit_record(); build_wetlab_hit_record() (enforces dry_lab_only=False); format_wetlab_hit_record()
• VALID_EXPERIMENT_TYPES (10): mic_broth_dilution, disk_diffusion, kill_curve, flow_cytometry_membrane, fluorescence_microscopy, biofilm_inhibition, hemolysis_assay, cytotoxicity_mtt, time_kill_assay, turbidity_growth_inhibition
• VALID_INTERPRETATIONS (3): active, inactive, inconclusive
• VALID_PROOF_LADDER_LEVELS: single_assay_hit, replicated_single_assay_hit, multi_assay_candidate
• tests/evidence/test_wetlab_hit_record.py: 59 tests in 5 classes
• BASELINE: 8648 → 8711

Why this matters

This is the FIRST schema where dry_lab_only must be False — it records actual wet-lab results, closing the computational nomination → experiment → result intake loop.

Disconfirming pass

• dry_lab_only=True raises violation (tested) — cannot pretend a real result is dry-lab
• TOY- candidate IDs blocked (tested) — real results require real candidate IDs
• Proof ladder is capped: no clinical efficacy claim allowed from a single assay result
• Empty limitations blocked (tested) — every result has scope limitations
• NaN/Inf/negative result_value blocked (tested)