| """Unit tests for the deterministic CounselBench criterion scorer.""" |
|
|
| from __future__ import annotations |
|
|
| import copy |
| import sys |
| import unittest |
| from pathlib import Path |
|
|
| RUNTIME = Path(__file__).resolve().parents[1] / "runtime" |
| if not RUNTIME.exists(): |
| RUNTIME = Path(__file__).resolve().parents[1] / "world" |
| sys.path.insert(0, str(RUNTIME)) |
|
|
| from scoring import aggregate_scores, score_findings, score_memo |
|
|
|
|
| def fixture() -> tuple[dict, dict, str]: |
| expected = { |
| "schema_version": "1.0", |
| "task_id": "cb100-test", |
| "matter_number": "CB-TEST-001", |
| "prepared_for": "Test Client, Inc.", |
| "as_of": "2026-09-04", |
| "findings": [ |
| { |
| "id": "F-01", |
| "issue": "consent deadline mismatch", |
| "severity": "high", |
| "primary_source": "/workspace/documents/primary.md", |
| "corroborating_source": "/workspace/documents/status.csv", |
| "determination": "The agreement requires consent by 2026-09-01; the status record remains pending on 2026-09-03.", |
| "recommended_action": "Deal Team must obtain consent by 2026-09-04.", |
| } |
| ], |
| } |
| scoring_row = { |
| **{key: expected["findings"][0][key] for key in ( |
| "id", "issue", "severity", "primary_source", "corroborating_source", |
| )}, |
| "fact_anchors": ["2026-09-01", "pending", "2026-09-03"], |
| "action_anchors": ["Deal Team", "obtain consent", "2026-09-04"], |
| "allowed_fact_text": "2026-09-01 pending 2026-09-03", |
| } |
| spec = { |
| "expected_findings": expected, |
| "scoring_findings": [scoring_row], |
| "memo_sections": [ |
| "Executive assessment", |
| "Method and record coverage", |
| "Findings", |
| "Recommended next actions", |
| "Assumptions and limitations", |
| ], |
| "forbidden_claims": ["not present in the production"], |
| } |
| memo = "\n".join( |
| [ |
| "# Executive assessment", |
| "## Method and record coverage", |
| "## Findings", |
| "F-01 consent deadline mismatch high /workspace/documents/primary.md ", |
| "/workspace/documents/status.csv 2026-09-01 pending 2026-09-03", |
| "## Recommended next actions", |
| "Deal Team obtain consent by 2026-09-04", |
| "## Assumptions and limitations", |
| ] |
| ) |
| return spec, expected, memo |
|
|
|
|
| class ScoringTests(unittest.TestCase): |
| def test_complete_grounded_outputs_pass(self) -> None: |
| spec, findings, memo = fixture() |
| finding_score = score_findings(findings, spec) |
| memo_score = score_memo(memo, spec) |
| procedure = {f"gate_{index}": True for index in range(8)} |
| procedure.update({ |
| "exact_deliverable_set": True, |
| "deliverables_written_through_mcp": True, |
| }) |
| aggregate = aggregate_scores(procedure, finding_score, memo_score) |
|
|
| self.assertEqual(len(finding_score["criteria"]), 17) |
| self.assertEqual(len(memo_score["criteria"]), 7) |
| self.assertTrue(aggregate["passed"]) |
| self.assertEqual(aggregate["reward"], 1.0) |
|
|
| def test_unsupported_controlled_fact_loses_only_its_criterion(self) -> None: |
| spec, findings, _ = fixture() |
| mutated = copy.deepcopy(findings) |
| mutated["findings"][0]["determination"] += " Exposure is $99,999,999." |
|
|
| result = score_findings(mutated, spec) |
|
|
| self.assertFalse(result["criteria"]["F-01.facts_source_bounded"]) |
| self.assertEqual(result["details"][0]["unsupported_fact_tokens"], ["$99,999,999"]) |
| self.assertGreater(result["score"], 0) |
| self.assertLess(result["score"], 1) |
|
|
| def test_wrong_source_path_does_not_erase_other_credit(self) -> None: |
| spec, findings, _ = fixture() |
| mutated = copy.deepcopy(findings) |
| mutated["findings"][0]["primary_source"] = "/workspace/documents/wrong.md" |
|
|
| result = score_findings(mutated, spec) |
|
|
| self.assertFalse(result["criteria"]["F-01.primary_source"]) |
| self.assertTrue(result["criteria"]["F-01.issue"]) |
| self.assertAlmostEqual(result["score"], 16 / 17, places=6) |
|
|
| def test_incomplete_procedure_caps_reward(self) -> None: |
| findings = {"score": 1.0, "passed": True} |
| memo = {"score": 1.0, "passed": True} |
| procedure = { |
| "review_complete": False, |
| "exact_deliverable_set": True, |
| "deliverables_written_through_mcp": True, |
| } |
|
|
| result = aggregate_scores(procedure, findings, memo) |
|
|
| self.assertEqual(result["reward"], 0.49) |
| self.assertEqual(result["cap_reason"], "required_review_procedure_incomplete") |
| self.assertFalse(result["passed"]) |
|
|
|
|
| if __name__ == "__main__": |
| unittest.main() |
|
|
