Close finite source review with explicit unresolved identity and income gaps

Reviewed-on: #46

apps/docs/scripts/auth-cli.ts

@@ -182,10 +182,10 @@ async function main() {
   assert.equal(researchResponse.status, 200, "authenticated research API");
   assert.ok(researchResponse.headers.get("cache-control")?.includes("no-store"));
   const research = await researchResponse.json();
-  assert.equal(research.companies.length, 1232);
+  assert.equal(research.companies.length, 1312);
   assert.equal(research.categories.length, 11);
-  assert.equal(research.evidence.length, 4797);
-  assert.equal(research.observations.length, 475);
+  assert.equal(research.evidence.length, 5354);
+  assert.equal(research.observations.length, 495);
   assert.equal((await request("/api/research/ai-companies", { method: "POST", headers })).status, 405);
   const researchPage = await request("/docs/other-research/ai-company-research/overview", { headers });
   assert.equal(researchPage.status, 200);

apps/docs/src/components/ArrDisclosures.tsx

@@ -34,7 +34,7 @@ export default function ArrDisclosures({ data, onSelect }: { data: ResearchSnaps
       <label>补证项目<select value={gap} onChange={(event) => setGap(event.target.value as ArrFieldGap | "")}><option value="">全部记录</option>{gapCounts.map((item) => <option key={item.id} value={item.id}>{gapLabels[item.id]}（{item.count} 条）</option>)}</select></label>
     </div>
     <div className="research-table-wrap"><table>
-      <caption>ARR 披露覆盖：按所选成立年份、每家公司计一次。点击分类筛选下方证据表。</caption>
+      <caption>ARR 披露覆盖：按公司当前分类及所选成立年份、每家公司计一次。点击分类筛选下方证据表。</caption>
       <thead><tr><th>业务分类</th><th>已采集公司</th><th>企业或当事方披露</th><th>仅第三方估算</th><th>缺少 ARR</th><th>披露覆盖率</th></tr></thead>
       <tbody>{coverage.filter((item) => item.companies > 0).map((item) => <tr key={item.id}>
         <td><button onClick={() => setCategory(item.id)} aria-pressed={category === item.id}>{item.name}</button></td>
@@ -42,7 +42,7 @@ export default function ArrDisclosures({ data, onSelect }: { data: ResearchSnaps
         <td>{new Intl.NumberFormat("zh-CN", { style: "percent", maximumFractionDigits: 1 }).format(item.disclosed / item.companies)}</td>
       </tr>)}</tbody>
     </table></div>
-    <p className="research-muted">分母含仅有目录资料的公司，覆盖率反映本次采集深度，不代表行业披露率。企业披露含历史里程碑，同一公司兼有披露和估算时归企业披露；缺少 ARR 的公司仍可能有其他收入。披露时点、币种与合同定义未对齐，不能将已披露少数公司的均值外推至整个分类，也不能由离散性判断财务造假。</p>
+    <p className="research-muted">分母含仅有目录资料的公司，覆盖率反映本次采集深度，不代表行业披露率。分类及收费模式采用公司当前资料；公司转向或同时经营多个产品时，历史 ARR 未必属于当前业务，需打开公司核对产品、主体与期间，不能据当前分类重归历史收入。企业披露含历史里程碑，同一公司兼有披露和估算时归企业披露；缺少 ARR 的公司仍可能有其他收入。披露时点、币种与合同定义未对齐，不能将已披露少数公司的均值外推至整个分类，也不能由离散性判断财务造假。</p>
     <aside className="research-notice" aria-label="ARR比较前的补证检查">
       <strong>ARR 比较前的补证检查</strong>
       <p>当前分类及成立年下有 {claims.length} 条披露，{completeFields} 条同时具有有效日期、明确 USD、精确金额且不是第三方估算。补证筛选后显示 {rows.length} 条。</p>
@@ -51,7 +51,7 @@ export default function ArrDisclosures({ data, onSelect }: { data: ResearchSnaps
     </aside>
     <div className="research-table-wrap"><table>
       <caption>金额保留原币种、下界和估算身份；日期栏保留业务截至日缺项</caption>
-      <thead><tr><th>公司 / 分类</th><th>收费与交付模式</th><th>原始 ARR 声明</th><th>时点</th><th>身份 / 限制</th><th>证据</th></tr></thead>
+      <thead><tr><th>公司 / 当前分类</th><th>当前收费与交付模式</th><th>原始 ARR 声明</th><th>时点</th><th>身份 / 限制</th><th>证据</th></tr></thead>
       <tbody>{rows.map(({ company, observation: item }) => {
         const source = evidence.get(item.evidence);
         const url = source && evidenceUrl(source.url);

apps/docs/tests/research.test.ts

@@ -92,7 +92,8 @@ test("ARR ledger preserves historical estimates without promoting targets, run r
   const ids = new Set(rows.map((row) => row.observation.id));
   for (const id of ["O064", "O065", "O066", "O069", "O070", "O071", "O046"])
     assert.ok(ids.has(id), `${id}: historical claims, estimates and unknown currencies remain visible`);
-  for (const id of ["O050", "O052", "O053", "O055", "SR-O-AI067-runrate", "ROCKET-O-arr", "AQ-O-RUNRATE"])
+  for (const id of ["O050", "O052", "O053", "O055", "SR-O-AI067-runrate", "ROCKET-O-arr", "AQ-O-RUNRATE",
+    "SRLEARN1-O-CLAIR-D2C", "SRLEARN1-O-CLAIR-B2B"])
     assert.ok(!ids.has(id), `${id}: excluded or ambiguous records cannot enter the ARR ledger`);
   assert.equal(rows.find((row) => row.observation.id === "O066")?.observation.selection, "historical-estimate");
   assert.equal(rows.find((row) => row.observation.id === "O046")?.observation.unit, "原文$，币种待核");
@@ -123,11 +124,11 @@ test("research revisions share the authentication database and preserve replay l
     for (let i = 0; i < 20; i++) consumeCode(config, "invalid");
     assert.equal(consumeCode(config, "invalid"), "limited");
     const initial = readResearchSnapshot();
-    assert.equal(initial.companies.length, 1232);
-    assert.equal(initial.evidence.length, 4797);
-    assert.equal(initial.observations.length, 475);
+    assert.equal(initial.companies.length, 1312);
+    assert.equal(initial.evidence.length, 5354);
+    assert.equal(initial.observations.length, 495);
     const businessReviews = initial.reviews.filter((review) => review.type === businessModelReviewType);
-    assert.equal(businessReviews.length, 6);
+    assert.equal(businessReviews.length, 26);
     assert.deepEqual(businessReviews, seed.reviews.filter((review) => review.type === businessModelReviewType)
       .sort((a, b) => a.id.localeCompare(b.id)),
       "business model assessments and company/evidence links persist in the shared database");
@@ -169,9 +170,9 @@ test("research validation rejects broken evidence and unsafe source links", () =
 
 test("reviews link explicit companies without turning global or excluded cases into companies", () => {
   const snapshot = structuredClone(seed) as ResearchSnapshot;
-  assert.equal(snapshot.reviews.filter((r) => r.scope === "companies").length, 234);
-  assert.equal(snapshot.reviews.filter((r) => r.scope === "dataset").length, 145);
-  assert.equal(snapshot.reviews.filter((r) => r.scope === "excluded-candidate").length, 53);
+  assert.equal(snapshot.reviews.filter((r) => r.scope === "companies").length, 299);
+  assert.equal(snapshot.reviews.filter((r) => r.scope === "dataset").length, 159);
+  assert.equal(snapshot.reviews.filter((r) => r.scope === "excluded-candidate").length, 65);
   const review = snapshot.reviews.find((r) => r.id === "R4")!;
   assert.deepEqual(reviewCompanyIds(snapshot, review), ["AI007", "AI013", "AI021", "AI034"]);
   assert.deepEqual(reviewCompanyIds(snapshot, snapshot.reviews.find((r) => r.id === "CN-R-PSI")!), []);
@@ -190,11 +191,11 @@ test("reviews link explicit companies without turning global or excluded cases i
 test("disclosure counts exclude client income, biographies, targets and duplicate milestones", () => {
   const snapshot = seed as ResearchSnapshot;
   const summary = categorySummary(snapshot);
-  assert.equal(summary.reduce((n, c) => n + c.count, 0), 1232);
-  assert.equal(summary.reduce((n, c) => n + c.self, 0), 62);
+  assert.equal(summary.reduce((n, c) => n + c.count, 0), 1312);
+  assert.equal(summary.reduce((n, c) => n + c.self, 0), 63);
   assert.equal(summary.reduce((n, c) => n + c.estimate, 0), 2);
   assert.equal(summary.reduce((n, c) => n + c.platform, 0), 19);
-  assert.equal(summary.reduce((n, c) => n + c.unknown, 0), 1149);
+  assert.equal(summary.reduce((n, c) => n + c.unknown, 0), 1228);
   assert.equal(snapshot.companies.find((c) => c.name === "Rork")?.revenueStatus, "unknown");
   assert.equal(snapshot.companies.find((c) => c.name === "Result")?.revenueStatus, "unknown");
   assert.ok(snapshot.observations.filter((o) => o.selection === "signal").every((o) => o.scope === "公司整体"));
@@ -229,8 +230,8 @@ test("service run rates, legacy freight revenue and paid adoption do not become
 
 test("technical reviews keep accessible materials separate from tested capabilities", () => {
   const snapshot = structuredClone(seed) as ResearchSnapshot;
-  assert.equal(snapshot.companies.filter((c) => c.technicalReview).length, 217);
-  assert.equal(snapshot.companies.filter((c) => c.technicalReview?.level === "public-artifact").length, 28);
+  assert.equal(snapshot.companies.filter((c) => c.technicalReview).length, 303);
+  assert.equal(snapshot.companies.filter((c) => c.technicalReview?.level === "public-artifact").length, 29);
   assert.equal(snapshot.companies.filter((c) => c.technicalReview?.level === "independent-test").length, 1);
   assert.equal(snapshot.companies.find((c) => c.id === "AI051")?.revenueStatus, "unknown");
   assert.match(snapshot.companies.find((c) => c.id === "AI051")!.technicalReview!.testScope!, /Sep 2026/);
@@ -567,6 +568,277 @@ test("the public SR007 roster preserves unique entries and Alven's older operato
     "annualized professional-service revenue is not recurring subscription ARR");
   assert.equal(snapshot.companies.find((item) => item.id === "AI137")!.technicalReview!.level, "description",
     "a preliminary external ranking with unresolved methodology is not complete independent verification");
+  const contractFollowup = snapshot.observations.filter((item) => item.id.startsWith("SR007F7-"));
+  assert.equal(contractFollowup.filter((item) => item.selection === "excluded").length, 4);
+  assert.equal(contractFollowup.filter((item) => item.selection === "adoption").length, 1);
+  assert.ok(contractFollowup.every((item) => !item.metric && !item.observedAt));
+  assert.ok(!arrDisclosures(snapshot).some(({ observation }) => observation.id.startsWith("SR007F7-")),
+    "contracted revenue and potential customers' active users cannot become realized ARR");
+  assert.equal(snapshot.companies.find((item) => item.id === "AI144")!.revenueStatus, "unknown",
+    "free current SDK terms and a contracted ARR claim need reconciliation before labeling realized revenue");
+  const scaledrive = batch.records.find((row) => row.name === "scaledrive.ai")!;
+  assert.equal(scaledrive.status, "outside-cohort", "SR007 membership cannot replace a stated 2024 founding year");
+  assert.equal(scaledrive.companyId, undefined);
+  assert.ok(scaledrive.evidenceIds?.includes("SR007F8-SCALEDRIVE-SR"));
+  for (const id of ["AI147", "AI150"]) {
+    assert.equal(snapshot.companies.find((item) => item.id === id)!.technicalReview!.level, "description",
+      "self-hosting and SDK documentation do not establish correctly attributed public source code");
+    assert.equal(snapshot.companies.find((item) => item.id === id)!.revenueStatus, "unknown");
+  }
+  assert.ok(!arrDisclosures(snapshot).some(({ observation }) => observation.id.startsWith("SR007F8-")),
+    "previous-company ARR and financial amounts in document demonstrations are not current revenue");
+  assert.equal(batch.records.filter((row) => row.status === "linked").length, 71);
+  assert.equal(batch.records.filter((row) => row.profileStatus === "archive-only").length, 0,
+    "all reused independent profiles have now been explicitly reread");
+  assert.equal(batch.records.filter((row) => row.profileStatus === "readable").length, 79);
+  const ninthFollowup = snapshot.observations.filter((item) => item.id.startsWith("SR007F9-"));
+  assert.equal(ninthFollowup.length, 10);
+  assert.equal(ninthFollowup.filter((item) => item.selection === "excluded").length, 6);
+  assert.equal(ninthFollowup.filter((item) => item.selection === "adoption").length, 3);
+  const industrialRunRate = ninthFollowup.find((item) => item.id === "SR007F9-O-VALENT-ANNUALIZED")!;
+  assert.equal(industrialRunRate.metric, "annualized-revenue");
+  assert.equal(industrialRunRate.selection, "historical");
+  assert.ok(ninthFollowup.every((item) => !item.observedAt),
+    "access dates and weeks since launch cannot supply precise financial dates");
+  assert.ok(ninthFollowup.filter((item) => item !== industrialRunRate).every((item) => !item.metric));
+  assert.ok(!arrDisclosures(snapshot).some(({ observation }) => observation.id.startsWith("SR007F9-")),
+    "annual contracts without fulfillment evidence, funding, annualized revenue and pipeline remain outside live ARR");
+  for (const id of ["AI152", "AI158"]) {
+    assert.equal(snapshot.companies.find((item) => item.id === id)!.revenueStatus, "unknown",
+      "contracted labels alone cannot establish realized revenue");
+  }
+  assert.equal(snapshot.companies.find((item) => item.id === "AI161")!.cat, "C02",
+    "enterprise workload model specialization is classified by delivery rather than frontier-model slogans");
+  const rereadProfiles = snapshot.evidence.filter((item) => item.id.startsWith("SR007F10-P-"));
+  assert.equal(rereadProfiles.length, 17);
+  assert.equal(new Set(rereadProfiles.map((item) => item.companyId)).size, 17);
+  for (const evidence of rereadProfiles) {
+    const candidate = batch.records.find((row) => row.companyId === evidence.companyId)!;
+    assert.equal(candidate.url, evidence.url);
+    assert.ok(candidate.evidenceIds?.includes(evidence.id), "each reread has an exact profile link");
+  }
+  assert.ok(!snapshot.observations.some((item) => item.id.startsWith("SR007F10-")),
+    "rereading existing financial milestones must not introduce duplicate observations");
+  for (const id of ["AI067", "AI069"]) {
+    assert.equal(snapshot.companies.find((item) => item.id === id)!.cat, "C02",
+      "current hosted-model and training-data delivery are infrastructure");
+  }
+  const oldWorkflowArr = snapshot.observations.find((item) => item.id === "SR-O-AI069-arr")!;
+  assert.equal(oldWorkflowArr.evidence, "SR-E-infragrid", "retain the original workflow revenue evidence");
+  assert.equal(oldWorkflowArr.value, 302000);
+  assert.equal(oldWorkflowArr.observedAt, undefined);
+  assert.ok(snapshot.companies.find((item) => item.id === "AI069")!.notes.includes("历史ARR归属不清"),
+    "a current category change must retain the unresolved historical product attribution");
+});
+
+test("public Speedrun cohort rosters preserve the finite boundary without inventing founding years", () => {
+  const snapshot = seed as ResearchSnapshot;
+  const rosters = snapshot.candidateBatches!.filter((batch) => /^SR00[1-7]-public-roster-2026-1$/.test(batch.id));
+  assert.equal(rosters.length, 7);
+  for (const [index, count] of [25, 25, 32, 42, 59, 59, 79].entries()) {
+    const batch = rosters.find((item) => item.id.startsWith(`SR${String(index + 1).padStart(3, "0")}-`))!;
+    assert.equal(batch.records.length, count);
+    assert.equal(batch.complete, false, "a complete directory page is not completed company-level research");
+  }
+  const rows = rosters.flatMap((batch) => batch.records);
+  assert.equal(rows.length, 321);
+  assert.equal(new Set(rows.map((row) => row.url)).size, 321);
+  const newlyRegistered = rosters.filter((batch) => !batch.id.startsWith("SR007-")).flatMap((batch) => batch.records);
+  assert.equal(newlyRegistered.length, 242);
+  assert.equal(newlyRegistered.filter((row) => row.status === "pending-year").length, 5);
+  assert.equal(newlyRegistered.filter((row) => row.status === "linked").length, 68);
+  assert.equal(newlyRegistered.filter((row) => row.status === "existing-outside-cohort").length, 1);
+  assert.equal(newlyRegistered.filter((row) => row.status === "outside-cohort").length, 144);
+  assert.equal(newlyRegistered.filter((row) => row.status === "pending-ai").length, 24,
+    "a known self-reported founding year does not by itself establish the current AI business");
+  assert.equal(newlyRegistered.filter((row) => row.profileStatus === "readable").length, 232);
+  assert.ok(newlyRegistered.filter((row) => row.status === "outside-cohort").every((row) =>
+    row.profileStatus === "readable" && row.evidenceIds?.some((id) => (id.startsWith("SRALL1-P-") || id.startsWith("SRPRO1-P-")))),
+  "old founding years require independent profile evidence, not cohort age");
+  const variants = newlyRegistered.filter((row) => row.name === "Variant");
+  assert.equal(variants.length, 2);
+  assert.equal(new Set(variants.map((row) => row.url)).size, 2, "same-name profiles must not be merged by name");
+  assert.equal(variants.filter((row) => row.companyId === "AI066").length, 1);
+  assert.ok(!snapshot.observations.some((row) => row.evidence.startsWith("SRALL1-")),
+    "roster discovery and outside-cohort claims do not create financial observations");
+  const profiles = snapshot.evidence.filter((row) => row.id.startsWith("SRPRO1-P-"));
+  assert.equal(profiles.length, 226);
+  for (const proof of profiles) {
+    const candidate = newlyRegistered.find((row) => row.id === proof.id.slice("SRPRO1-P-".length))!;
+    assert.equal(candidate.url, proof.url);
+    assert.ok(candidate.evidenceIds?.includes(proof.id));
+  }
+  for (const id of ["AI162", "AI163", "AI164", "AI165", "AI166", "AI167"]) {
+    const company = snapshot.companies.find((row) => row.id === id)!;
+    assert.equal(company.year, 2025);
+    assert.equal(company.revenueStatus, "unknown");
+    assert.equal(company.technicalReview!.level, "description");
+  }
+  assert.ok(!snapshot.observations.some((row) => row.evidence.startsWith("SRPRO1-")),
+    "future GPU capacity, free trials, customer savings and unread financial claims are not realized revenue");
+});
+
+test("Speedrun brand dates and survey benchmark scores do not create new-company revenue", () => {
+  const snapshot = seed as ResearchSnapshot;
+  const rows = snapshot.candidateBatches!.flatMap((batch) => batch.records);
+  for (const id of ["SR004R-C11", "SR004R-C20"]) {
+    const row = rows.find((item) => item.id === id)!;
+    assert.equal(row.status, "pending-year", "a known brand year must retain unresolved underlying entity dates");
+    assert.equal(row.companyId, undefined);
+    assert.ok(row.evidenceIds!.some((proof) => proof.startsWith("SR004F2-")));
+  }
+  for (const id of ["SR004R-C08", "SR004R-C15"]) {
+    assert.equal(rows.find((item) => item.id === id)!.status, "pending-ai");
+  }
+  for (const [id, category] of [["AI168", "C06"], ["AI169", "C05"]]) {
+    const company = snapshot.companies.find((item) => item.id === id)!;
+    assert.equal(company.cat, category);
+    assert.equal(company.revenueStatus, "unknown");
+    assert.equal(company.technicalReview!.level, "description");
+  }
+  assert.ok(!snapshot.observations.some((row) => row.evidence.startsWith("SR004F2-")),
+    "brand-deal opportunities, listed fees and survey agreement are not financial observations");
+  const model = snapshot.reviews.find((row) => row.id === "MODEL1-SIMULATION-PANEL")!;
+  assert.deepEqual(reviewCompanyIds(snapshot, model), ["AI169"]);
+  assert.ok(model.evidenceIds.includes("SR004F2-REHEARSALS-PRICE"));
+});
+
+test("recruiting service adoption and official balance sheets do not become ARR", () => {
+  const snapshot = seed as ResearchSnapshot;
+  for (const id of ["AI170", "AI171", "AI172", "AI173", "AI174", "AI175"]) {
+    const company = snapshot.companies.find((row) => row.id === id)!;
+    assert.equal(company.year, 2025);
+    assert.equal(company.revenueStatus, "unknown");
+    assert.equal(company.technicalReview!.level, "description");
+  }
+  const accounts = snapshot.evidence.find((row) => row.id === "SR005F3-DEX-ACCOUNTS")!;
+  assert.equal(accounts.companyId, "AI171");
+  assert.equal(new URL(accounts.url).hostname, "find-and-update.company-information.service.gov.uk");
+  assert.equal(accounts.publishedAt, "2026-09-08", "filing date is preserved without becoming a revenue measurement date");
+  assert.ok(!snapshot.observations.some((row) => row.evidence.startsWith("SR005F3-")),
+    "balance-sheet cash, client quotes, salaries and recruiting outcomes do not disclose revenue");
+  const model = snapshot.reviews.find((row) => row.id === "MODEL1-RECRUITING-SERVICE")!;
+  assert.deepEqual(reviewCompanyIds(snapshot, model), ["AI175", "AI171"]);
+  assert.ok(model.evidenceIds.includes("SR005F3-JOOBA-TERMS"));
+});
+
+test("current products preserve unresolved identity and separate trading from software revenue", () => {
+  const snapshot = seed as ResearchSnapshot;
+  const presia = snapshot.candidateBatches!.flatMap((batch) => batch.records)
+    .find((row) => row.id === "SR005R-C44")!;
+  assert.equal(presia.status, "pending-ai");
+  assert.equal(presia.companyId, undefined, "a reused domain does not resolve a changed business entity");
+  assert.ok(presia.evidenceIds!.includes("SR005F4-PRESIA-ASSET"));
+  for (const id of ["AI176", "AI177", "AI178", "AI179", "AI180"]) {
+    const company = snapshot.companies.find((row) => row.id === id)!;
+    assert.equal(company.revenueStatus, "unknown");
+  }
+  const syzygy = snapshot.companies.find((row) => row.id === "AI180")!;
+  assert.equal(syzygy.technicalReview!.level, "public-artifact");
+  assert.ok(syzygy.technicalReview!.evidenceIds.includes("SR005F4-SYZYGY-FILES"));
+  assert.ok(!snapshot.observations.some((row) => row.evidence.startsWith("SR005F4-")),
+    "landed prices, customer savings, model scores and listed subscriptions are not reported ARR");
+  const supply = snapshot.reviews.find((row) => row.id === "MODEL1-PHYSICAL-SUPPLY")!;
+  assert.deepEqual(reviewCompanyIds(snapshot, supply), ["AI179"]);
+  assert.ok(snapshot.reviews.find((row) => row.id === "MODEL1-LOCAL-CLOUD")!.companyIds!.includes("AI180"));
+});
+
+test("new founding-year reports retain beta, academic and self-reported enforcement boundaries", () => {
+  const snapshot = seed as ResearchSnapshot;
+  for (const id of ["AI181", "AI182", "AI184", "AI185", "AI186"]) {
+    const company = snapshot.companies.find((row) => row.id === id)!;
+    assert.equal(company.year, 2026);
+    assert.equal(company.revenueStatus, "unknown");
+    assert.equal(company.technicalReview!.level, "description");
+  }
+  assert.equal(snapshot.companies.find((row) => row.id === "AI183")!.year, 2025);
+  assert.ok(!snapshot.observations.some((row) => row.evidence.startsWith("SR006F5-")),
+    "trial credits, customer revenue impact, demonstration fundraises and normalized research scores are not ARR");
+  const research = snapshot.reviews.find((row) => row.id === "SR006F5-R-RESEARCH")!;
+  assert.deepEqual(reviewCompanyIds(snapshot, research), ["AI185", "AI169"]);
+  assert.ok(research.evidenceIds.includes("SR004F2-REHEARSALS-BENCH"));
+  const pricing = snapshot.reviews.find((row) => row.id === "MODEL1-MESSAGE-ENFORCEMENT")!;
+  assert.deepEqual(reviewCompanyIds(snapshot, pricing), ["AI186"]);
+  assert.ok(pricing.evidenceIds.includes("SR006F5-ZERODRIFT-TERMS"));
+});
+
+test("rebrands and ambiguous monthly receipts stay outside comparable ARR", () => {
+  const snapshot = seed as ResearchSnapshot;
+  const rows = snapshot.candidateBatches!.flatMap((batch) => batch.records);
+  for (const id of ["SR006R-C37", "SR006R-C54"]) {
+    const row = rows.find((item) => item.id === id)!;
+    assert.equal(row.status, "pending-year");
+    assert.equal(row.companyId, undefined, "brand dates must not overwrite existing legal-entity dates");
+  }
+  const idilio = rows.find((item) => item.id === "SR006R-C22")!;
+  assert.equal(idilio.status, "pending-ai");
+  assert.ok(idilio.evidenceIds!.includes("SR006F6-IDILIO-SR"));
+  assert.ok(!snapshot.observations.some((row) => row.evidence.startsWith("SR006F6-")),
+    "pipeline, retained client revenue, illustrative tax values and undated monthly receipts are not comparable ARR");
+  for (const id of ["AI187", "AI188", "AI189"]) {
+    assert.equal(snapshot.companies.find((row) => row.id === id)!.revenueStatus, "unknown");
+  }
+  const tax = snapshot.reviews.find((row) => row.id === "MODEL1-TAX-PREP")!;
+  assert.deepEqual(reviewCompanyIds(snapshot, tax), ["AI188"]);
+  assert.ok(tax.evidenceIds.includes("SR006F6-GROVE-PRICE"));
+});
+
+test("care and legal delivery preserve financial exclusions and linked cost reviews", () => {
+  const snapshot = seed as ResearchSnapshot;
+  for (const id of ["AI190", "AI191", "AI192", "AI193", "AI194", "AI195"]) {
+    const company = snapshot.companies.find((row) => row.id === id)!;
+    assert.equal(company.revenueStatus, "unknown");
+    assert.equal(company.technicalReview!.level, "description");
+    const candidate = snapshot.candidateBatches!.flatMap((batch) => batch.records)
+      .find((row) => row.companyId === id)!;
+    assert.equal(candidate.status, "linked");
+    assert.equal(candidate.categoryId, company.cat);
+    assert.ok(company.evidenceIds.every((proof) => snapshot.evidence.some((row) => row.id === proof && row.companyId === id)));
+  }
+  assert.ok(!snapshot.observations.some((row) => row.evidence.startsWith("SRCARE1-")),
+    "undefined annual run rate, customer savings, list prices and usage are not comparable ARR");
+  const hardware = snapshot.reviews.find((row) => row.id === "MODEL1-HARDWARE-LIFETIME")!;
+  assert.deepEqual(reviewCompanyIds(snapshot, hardware), ["AI195"]);
+  const clinical = snapshot.reviews.find((row) => row.id === "MODEL1-CARE-OPERATIONS")!;
+  assert.deepEqual(reviewCompanyIds(snapshot, clinical), ["AI191", "AI194"]);
+});
+
+test("enterprise contracts and launch-year conflicts retain their exclusions", () => {
+  const snapshot = structuredClone(seed) as ResearchSnapshot;
+  const accelerator = snapshot.candidateBatches!.flatMap((batch) => batch.records)
+    .find((row) => row.id === "SR006R-C01")!;
+  assert.equal(accelerator.status, "pending-year");
+  assert.equal(accelerator.companyId, undefined);
+  assert.ok(accelerator.evidenceIds!.includes("SRENTER1-ACCELER8-DEALROOM"));
+  for (const [id, value, bound] of [["HEAVI", 100000, "="], ["VERRA", 700000, "≈"]] as const) {
+    const contract = snapshot.observations.find((row) => row.id === `SRENTER1-O-${id}-CONTRACT`)!;
+    assert.equal(contract.value, value);
+    assert.equal(contract.bound, bound);
+    assert.equal(contract.selection, "excluded");
+    assert.equal(contract.observedAt, undefined);
+    assert.equal(contract.comparisonGroup, null);
+    assert.equal(snapshot.companies.find((row) => row.id === contract.company)!.revenueStatus, "unknown");
+    contract.metric = "arr";
+    assert.ok(!arrDisclosures(snapshot).some((row) => row.observation.id === contract.id),
+      "even a wrongly added ARR label must not promote an excluded contract");
+    assert.ok(!platformIncomeSample(snapshot, { windowDays: 90 }).points
+      .some((row) => row.observation.id === contract.id));
+  }
+});
+
+test("financial candidates need current AI evidence before they acquire company and income links", () => {
+  const snapshot = seed as ResearchSnapshot;
+  const candidates = snapshot.candidateBatches!.flatMap((batch) => batch.records);
+  for (const id of ["SR006R-C06", "SR006R-C16", "SR005R-C35", "SR006R-C08"]) {
+    const candidate = candidates.find((row) => row.id === id)!;
+    assert.equal(candidate.status, "pending-ai");
+    assert.equal(candidate.companyId, undefined);
+    assert.ok(candidate.evidenceIds!.some((proof) => proof.startsWith("SRFIN1-")));
+  }
+  const financial = snapshot.reviews.find((row) => row.id === "MODEL1-FINANCIAL-OPERATIONS")!;
+  assert.deepEqual(reviewCompanyIds(snapshot, financial), ["AI202", "AI203", "AI204"]);
+  assert.ok(!snapshot.observations.some((row) => row.evidence.startsWith("SRFIN1-")),
+    "bank size, FX processing, artist revenue, demo portfolios and benchmark scores are not vendor income");
 });
 
 test("annual launch archives exclude older companies only with evidence and reuse existing records", () => {
@@ -596,7 +868,7 @@ test("unreconciled directory financials remain source reviews instead of income
   assert.ok(review.evidenceIds.includes("PH-TANA-DR"));
   assert.equal(reviewCompanyIds(snapshot, review).length, 0);
   assert.ok(!snapshot.observations.some((o) => o.evidence === "PH-TANA-DR"));
-  assert.equal(snapshot.candidateBatches!.flatMap((b) => b.records).filter((r) => r.status === "pending-year").length, 153);
+  assert.equal(snapshot.candidateBatches!.flatMap((b) => b.records).filter((r) => r.status === "pending-year").length, 136);
 });
 
 test("product pivots and founder histories do not become new company years or current revenue", () => {

content/research/ai-income-analysis.json

@@ -1,5 +1,5 @@
 {
-  "sourceVersion": "2026-11-01-sr007-followup-6",
+  "sourceVersion": "2026-11-01-source-closeout-1",
   "asOf": "2026-11-01",
   "metric": "platform-mrr-usd",
   "sampleUnit": "product-or-business",

scripts/audit-ai-research.py

@@ -4,7 +4,7 @@ from collections import Counter
 import json
 from pathlib import Path
 import sys
-from urllib.parse import urlsplit
+from urllib.parse import parse_qs, urlsplit
 
 ROOT = Path(__file__).resolve().parents[1]
 DATA = json.loads((ROOT / 'content/research/ai-companies.seed.json').read_text())
@@ -49,6 +49,54 @@ def source_register():
             assert roster['defaultCohort'] == 'SR007' and roster['next'] is None
             assert len(batch['records']) == roster['recordCount'] == roster['reportedCount']
             assert len({row['url'] for row in batch['records']}) == roster['recordCount']
+        if source['id'] == 'S14' and source.get('publicRosterAudits'):
+            audits = source['publicRosterAudits']
+            assert {a['cohort'] for a in audits} == {f'SR{i:03d}' for i in range(1, 8)}
+            assert len(audits) == 7
+            profile_urls = []
+            for audit in audits:
+                batch = next(b for b in DATA['candidateBatches'] if b['id'] == audit['candidateBatchId'])
+                query = parse_qs(urlsplit(audit['url']).query)
+                assert urlsplit(audit['url']).netloc == 'speedrun.a16z.com'
+                assert query == {'limit': ['100'], 'offset': ['0'], 'cohort': [audit['cohort']]}
+                assert urlsplit(audit['url']).path == '/api/companies/companyparams'
+                assert audit['httpStatus'] == 200 and audit['accessedAt'] <= DATA['asOf']
+                assert audit['next'] is None and audit['previous'] is None
+                assert len(batch['records']) == audit['recordCount'] == audit['reportedCount']
+                assert audit['uniquePublicIds'] == audit['uniqueSlugs'] == audit['recordCount']
+                assert len(audit['responseSha256']) == 64 and all(c in '0123456789abcdef' for c in audit['responseSha256'])
+                assert batch['sourceId'] == 'S14' and not batch['complete']
+                proof = next(e for e in DATA['evidence'] if e['id'] == audit['evidenceId'])
+                assert proof['url'] == audit['url'] and proof['companyId'] is None
+                profile_urls.extend(row['url'] for row in batch['records'])
+            assert len(profile_urls) == len(set(profile_urls)), 'duplicate profiles across public cohorts'
+        if source['id'] == 'S14' and source.get('profileAudit'):
+            audit = source['profileAudit']
+            entries = audit['records']
+            candidates = {row['id']: row for b in DATA['candidateBatches'] for row in b['records']}
+            evidence_by_id = {row['id']: row for row in DATA['evidence']}
+            assert len(entries) == audit['recordCount'] == audit['readable']
+            assert audit['failed'] == 0 and len({e['candidateId'] for e in entries}) == len(entries)
+            years = Counter(str(e['foundedYear']) for e in entries)
+            assert dict(years) == audit['yearCounts']
+            for entry in entries:
+                row = candidates[entry['candidateId']]
+                proof = evidence_by_id[entry['evidenceId']]
+                assert row['profileStatus'] == 'readable' and proof['url'] == row['url']
+                assert entry['evidenceId'] in row['evidenceIds']
+                assert isinstance(entry['foundedYear'], int) and 1800 <= entry['foundedYear'] <= int(DATA['asOf'][:4])
+                assert isinstance(entry['teamSize'], int) and entry['teamSize'] >= 0
+                assert len(entry['htmlSha256']) == 64 and all(c in '0123456789abcdef' for c in entry['htmlSha256'])
+                if entry['foundedYear'] not in [2025, 2026]:
+                    assert row['status'] == 'outside-cohort' and not row.get('companyId')
+            assert sum(e['foundedYear'] not in [2025, 2026] for e in entries) == audit['outsideTargetYears']
+            assert sum(e['foundedYear'] in [2025, 2026] for e in entries) == audit['targetYearSelfReports']
+            assert sum(candidates[e['candidateId']]['status'] == 'linked' for e in entries) == audit['newLinkedCompanies']
+            assert sum(candidates[e['candidateId']]['status'] == 'pending-ai' for e in entries) == audit['pendingCurrentAIOrIdentity']
+            assert sum(candidates[e['candidateId']]['status'] == 'pending-year' for e in entries) == audit['pendingFoundingIdentity']
+            assert audit['targetYearSelfReports'] == (
+                audit['newLinkedCompanies'] + audit['pendingCurrentAIOrIdentity'] + audit['pendingFoundingIdentity']
+            ), 'self-reported years must retain unresolved brand/entity dates'
         if source['url']:
             assert urlsplit(source['url']).scheme in {'http', 'https'} and urlsplit(source['url']).netloc
         for host in source['hosts']: