Merge pull request '完成 SR007 名单逐项档案读取，补充二十一家公司并区分历史收入归属' (#44) from codex/ai-research-sr007-more into main

Reviewed-on: #44

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, 1253);
   assert.equal(research.categories.length, 11);
-  assert.equal(research.evidence.length, 4797);
-  assert.equal(research.observations.length, 475);
+  assert.equal(research.evidence.length, 4894);
+  assert.equal(research.observations.length, 491);
   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

@@ -123,11 +123,11 @@ 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, 1253);
+    assert.equal(initial.evidence.length, 4894);
+    assert.equal(initial.observations.length, 491);
     const businessReviews = initial.reviews.filter((review) => review.type === businessModelReviewType);
-    assert.equal(businessReviews.length, 6);
+    assert.equal(businessReviews.length, 7);
     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 +169,9 @@ 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, 248);
+  assert.equal(snapshot.reviews.filter((r) => r.scope === "dataset").length, 147);
+  assert.equal(snapshot.reviews.filter((r) => r.scope === "excluded-candidate").length, 54);
   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 +190,11 @@ 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), 1253);
+  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), 1169);
   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"));
@@ -229,7 +229,7 @@ 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).length, 244);
   assert.equal(snapshot.companies.filter((c) => c.technicalReview?.level === "public-artifact").length, 28);
   assert.equal(snapshot.companies.filter((c) => c.technicalReview?.level === "independent-test").length, 1);
   assert.equal(snapshot.companies.find((c) => c.id === "AI051")?.revenueStatus, "unknown");
@@ -567,6 +567,67 @@ 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("annual launch archives exclude older companies only with evidence and reuse existing records", () => {
@@ -596,7 +657,7 @@ 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, 131);
 });
 
 test("product pivots and founder histories do not become new company years or current revenue", () => {

content/research/ai-companies.sources.json

@@ -1,5 +1,5 @@
 {
-  "version": "2026-11-01-sr007-followup-6",
+  "version": "2026-11-01-sr007-followup-10",
   "scopeConfirmedAt": "2026-10-30",
   "acceptance": "以确认的公开来源清单为边界，逐条核对并标明缺项。",
   "provenance": "S01—S12恢复自已提交用户的《AI公司研究_来源与分类模板》来源清单；S13—S16单列后续已使用的补充证据渠道，不将使用次数当作全量覆盖。",
@@ -3040,8 +3040,8 @@
         "speedrun.a16z.com"
       ],
       "initialAccess": "见具体证据访问记录",
-      "boundary": "原有定向个案保留。2026-11-01匿名普通HTML默认SR007名单79条，75个普通UUID链接；逐一登记50关联（17复用、33本系列新增）、23成立/主体待核、6AI主营待核。本系列补读40个独立档案及可读官网/条款，剩余目录位置未当独立档案已读。未执行站点脚本或读取私有接口；SR005/SR006参数仍返回SR007，不认其他批次或两年全量覆盖。Bric/Coherence/KIRA/Leonidas/Mecone/PhotonSpear当前AI主营未明，Alven有2024主体与2025品牌冲突。签约/管线/预计ARR、客户资产、旧事业、候补与演示不当实收；Maestro九个月累计收入与Afore的两个ARR里程碑分列，后者缺币种/日历日及净经常性定义。软件、人工/专家、云资源、评测、硬件/旅行交易及保险安排按当前交付分开，技术制品与独立验证分开；具体限制链接公司、证据和核查记录。 OpenWhispr新增有缺项ARR标签与日活采用，本地/BYOK/自营云分开，旧Hedra/StockX/金融产品业绩及客户运费不迁新公司。 Othello live/cARR/管线、Parca服务年化、Phyvant客户财务及Orbits当前免费与旧雇主业绩分别保留；Oruk外部英语初步榜单不能外推整体效果。",
-      "nextCheck": "先补SR007剩余23条独立档案及Bric/Coherence/KIRA/Leonidas/Mecone/PhotonSpear的AI主营，另核Alven主体/品牌沿革。继续建立其他批次公开边界，不能按请求参数认读取成功。已关联公司仍须补独立登记、币种/财务日、签约履约/实付、经常性净费及取消退款；实施/人工、数据许可、评测、云/模型资源、设备、保险佣金/保费与旅行采购及会员费需拆。Hoid性能/合同、Kairo工程师服务、Keythorn牌照/风险角色、Maestro2025推荐语/主体起点及两种收入口径为当前重点；其他个案缺项见关联核查，不因本次可读认已验证。 Mirai业务订单与模型依赖、Motherboard现场人力/装机、Munari评分/专家成本、Nimble客户审批/付费、OpenWhispr免费与云净费/隐私配置和独立主体继续核。 Oruk精确模型/方法与语音用量成本，Othello主体/品牌起点及live合同净额，Parca专业审核净费/资质，Orbits免费承诺/未来定价及Phyvant实施/异常人工与合同继续缺。",
+      "boundary": "原有定向个案保留。2026-11-01匿名普通HTML默认SR007载荷79条唯一ID/slug、75普通UUID链接，count79且next/previous空。当前71关联（17复用、54本系列新增）、Alven1年份/主体冲突、6AI主营待核、scaledrive2024超年1；本系列已读79独立档案（62补读＋17复用重读），全部位置有对应档案证据；复用原ID/观测，不新增重复公司。未执行站点脚本/私有接口；SR005/SR006参数仍返回SR007，不认其他批次或两年全量。Bric/Coherence/KIRA/Leonidas/Mecone/PhotonSpear当前AI主营未明。签约/cARR/管线/融资、客户资产/回款、旧事业业绩和界面财务不当已实现收入；live ARR、年度合同标签、年化服务和累计收入分别保留，财务日期/币种/净费不补造。软件订阅、额度充值、模型定制/运行、专业/人工托管、云存储、评测、硬件、金融手续费及旅行采购按交付拆开。所有公司交付/技术/收费缺项见关联证据和核查记录；公开说明、未发布功能、公开制品与独立测试分级。Sterling条款法人占位及Zoah的OPACITY残留只作补证，不能据此推造假。 本轮官网复核中Infragrid首页实验室数据/评测与条款Runbooks并列，当前暂C02但旧ARR不迁产品；Abliteration托管/API按C02，标题/正文口径不明仍保留。Codos/Proper明确含嵌入式工程，非纯自助软件假定。",
+      "nextCheck": "已完成17复用独立档案重读；继续核Alven2024主体/2025品牌沿革及6项AI主营。再建立其他批次可验证公开边界，不按请求参数认成功。公司层面补独立登记、真实合同/履约/取消、币种与财务日、经常性净费、模型/人工/资料授权/存储和渠道成本；重点包括Sterling年合同净费与法人、Vene个人/机构实付、Valent年化业务组成、Truli单次/订阅、Zoah训练退出/席位额度/运营者、Foundry公告/制品及Vound具体产品。其余公司核查缺项继续保留，不以本批可读证明全部真实。 Infragrid数据业务/旧Runbooks的产品沿革、合同与数据许可，Abliteration衍生模型权利/预收消费，Alia Prosieve及Proper GAIA/Virgo主体关系，Codos/Proper试点转续费及工程工时需补。",
       "limitation": "来源可读不等于声明已独立验证。",
       "rosterAudit": {
         "url": "https://speedrun.a16z.com/companies",
@@ -3083,8 +3083,8 @@
         "benchmarks.speko.ai"
       ],
       "initialAccess": "见具体证据访问记录",
-      "boundary": "未归入上述来源的公开材料按原始证据链接逐条保留；不视为新增已批准的完整候选池。 本轮读取GetLatka单页历史年收入估算，方法底表未取得，不视为新增完整候选池。 另定向读GuideAI估算页，明确EST且团队身份与当前原始线索未对齐，本批不采财务观测；非全库。 定向读取Afore公开Maestro公司页（同名创始人及官网反链），记录ARR原始标签但无日历截至日/币种或净费定义；未将其Showcase全名单视为已批准候选池。 定向读取Speko英语情绪初步榜单，仅核Oruk公开结果，方法/独立性/版本待核，非扩展整个候选池。",
-      "nextCheck": "逐项核来源归属、可靠性及可读正文，不能将搜索摘要当全文。 Maestro投资方ARR与累计收入、净会员费/旅行总额及顾问成本需对齐；不请求介绍。",
+      "boundary": "未归入上述来源的公开材料按原始证据链接逐条保留；不视为新增已批准的完整候选池。 本轮读取GetLatka单页历史年收入估算，方法底表未取得，不视为新增完整候选池。 另定向读GuideAI估算页，明确EST且团队身份与当前原始线索未对齐，本批不采财务观测；非全库。 定向读取Afore公开Maestro公司页（同名创始人及官网反链），记录ARR原始标签但无日历截至日/币种或净费定义；未将其Showcase全名单视为已批准候选池。 定向读取Speko英语情绪初步榜单，仅核Oruk公开结果，方法/独立性/版本待核，非扩展整个候选池。 定向读取Afore公开Showcase中Vene同三名创始人的家庭采用、contracted ARR与管线；独立公司页超时未读，不当新增全池或已实现ARR。",
+      "nextCheck": "逐项核来源归属、可靠性及可读正文，不能将搜索摘要当全文。 Maestro投资方ARR与累计收入、净会员费/旅行总额及顾问成本需对齐；不请求介绍。 Vene合同履约、个人/赞助付费与家庭分母需补，禁止以家庭×价格估计已实现ARR。",
       "limitation": "来源可读不等于声明已独立验证。"
     }
   ]

content/research/ai-income-analysis.json

@@ -1,5 +1,5 @@
 {
-  "sourceVersion": "2026-11-01-sr007-followup-6",
+  "sourceVersion": "2026-11-01-sr007-followup-10",
   "asOf": "2026-11-01",
   "metric": "platform-mrr-usd",
   "sampleUnit": "product-or-business",