docs(rectification): record v3 sealed holdout as evaluated not missing

Co-authored-by: Cursor <cursoragent@cursor.com>
This commit is contained in:
Jesse_Chen
2026-08-29 00:25:26 +08:00
parent fe0afdaddb
commit 9d80498316
4 changed files with 60 additions and 20 deletions
@@ -33,11 +33,17 @@ import {
const THIRD_CANDIDATE_ID = "88888888-8888-4888-8888-888888888883";
const holdoutReport = JSON.parse(readFileSync(
new URL("../../references/real_case_calibration/minute_rectification_holdout_v2_pilot_report.json", import.meta.url),
new URL("../../references/real_case_calibration/minute_rectification_holdout_v3_report.json", import.meta.url),
"utf8",
)) as {
case_gate: { valid_public_aa_cases: number; minimum_public_aa_cases: number };
metrics: { top_1_rate: number; confirmation_coverage_rate: number };
benchmark_id: string;
validation: { valid_public_aa_cases: number; minimum_public_aa_cases: number };
release_metrics: {
top_1_rate: { actual: number };
top_3_rate: { actual: number };
mean_absolute_minute_error: { actual: number };
confirmation_coverage_rate: number;
};
};
const skill = readFileSync(
new URL("../../skills/jyotish-birth-time-rectification/SKILL.md", import.meta.url),
@@ -80,19 +86,33 @@ function blocker(
return found;
}
test("sealed holdout aggregates match the v2 report and stay below the case gate", () => {
test("sealed holdout aggregates match the v3 report and still block confirmation", () => {
const productHoldout = JSON.parse(readFileSync(
new URL("../../references/rectification_sealed_holdout.v1.json", import.meta.url),
"utf8",
)) as typeof SEALED_MINUTE_HOLDOUT;
assert.equal(SEALED_MINUTE_HOLDOUT.valid_public_aa_cases, holdoutReport.case_gate.valid_public_aa_cases);
assert.equal(SEALED_MINUTE_HOLDOUT.required_cases, holdoutReport.case_gate.minimum_public_aa_cases);
assert.equal(SEALED_MINUTE_HOLDOUT.top_1_rate, holdoutReport.metrics.top_1_rate);
assert.equal(SEALED_MINUTE_HOLDOUT.confirmation_coverage_rate, holdoutReport.metrics.confirmation_coverage_rate);
assert.equal(SEALED_MINUTE_HOLDOUT.sealed_benchmark_id, holdoutReport.benchmark_id);
assert.equal(SEALED_MINUTE_HOLDOUT.valid_public_aa_cases, holdoutReport.validation.valid_public_aa_cases);
assert.equal(SEALED_MINUTE_HOLDOUT.required_cases, holdoutReport.validation.minimum_public_aa_cases);
assert.equal(SEALED_MINUTE_HOLDOUT.top_1_rate, holdoutReport.release_metrics.top_1_rate.actual);
assert.equal(
SEALED_MINUTE_HOLDOUT.confirmation_coverage_rate,
holdoutReport.release_metrics.confirmation_coverage_rate,
);
assert.equal(SEALED_MINUTE_HOLDOUT.valid_public_aa_cases, productHoldout.valid_public_aa_cases);
assert.equal(SEALED_MINUTE_HOLDOUT.status, productHoldout.status);
assert.ok(SEALED_MINUTE_HOLDOUT.valid_public_aa_cases < SEALED_MINUTE_HOLDOUT.required_cases);
assert.ok(SEALED_MINUTE_HOLDOUT.valid_public_aa_cases >= SEALED_MINUTE_HOLDOUT.required_cases);
assert.equal(SEALED_MINUTE_HOLDOUT.status, "not_ready");
assert.equal(SEALED_MINUTE_HOLDOUT.confirmation_coverage_rate, 0);
const gate = buildConfirmationGate({
engineConfirmationAllowed: true,
candidates: UNIQUE_MINUTE,
decisionReceipt: {
gates: { exact_confirmation: { external_validation_status: "passed" } },
},
});
assert.equal(gate.confirmation_allowed, false);
assert.equal(blocker(gate, "public_aa_holdout").status, "not_ready");
});
test("VedAstro not_evaluated is not described as fail and still forbids confirmation", () => {