merge: include latest company research updates from PR #23

2026-10-30 18:20:34 +08:00
parent 083f532609 770be83692
commit 15024c3908
5 changed files with 224 additions and 14 deletions
@@ -184,7 +184,7 @@ async function main() {
  const research = await researchResponse.json();
  assert.equal(research.companies.length, 1023);
  assert.equal(research.categories.length, 11);
-  assert.equal(research.evidence.length, 1251);
+  assert.equal(research.evidence.length, 1259);
  assert.equal(research.observations.length, 189);
  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 });
@@ -24,7 +24,7 @@ test("research revisions share the authentication database and preserve replay l
    assert.equal(consumeCode(config, "invalid"), "limited");
    const initial = readResearchSnapshot();
    assert.equal(initial.companies.length, 1023);
-    assert.equal(initial.evidence.length, 1251);
+    assert.equal(initial.evidence.length, 1259);
    assert.equal(initial.observations.length, 189);
    assert.deepEqual(initial.candidateBatches, seed.candidateBatches, "candidate gaps persist in the same database metadata");
    assert.deepEqual(initial.directoryRechecks, seed.directoryRechecks, "follow-up classifications persist without rewriting the discovery audit");
@@ -64,9 +64,9 @@ 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, 59);
-  assert.equal(snapshot.reviews.filter((r) => r.scope === "dataset").length, 8);
-  assert.equal(snapshot.reviews.filter((r) => r.scope === "excluded-candidate").length, 28);
+  assert.equal(snapshot.reviews.filter((r) => r.scope === "companies").length, 60);
+  assert.equal(snapshot.reviews.filter((r) => r.scope === "dataset").length, 9);
+  assert.equal(snapshot.reviews.filter((r) => r.scope === "excluded-candidate").length, 29);
  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")!), []);
@@ -307,7 +307,7 @@ 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, 5);
+  assert.equal(snapshot.candidateBatches!.flatMap((b) => b.records).filter((r) => r.status === "pending-year").length, 8);
 });

 test("product pivots and founder histories do not become new company years or current revenue", () => {
@@ -1,5 +1,5 @@
 {
-  "version": "2026-10-30-thrive-brila-1",
+  "version": "2026-10-30-archives-1",
  "asOf": "2026-10-30",
  "scope": "全球公开资料研究：保留首批个案，并扩展核查2025—2026八个YC批次候选池；成立年份以官方公司页面字段为依据。",
  "limitations": [
@@ -10387,7 +10387,7 @@
      "id": "YC30879",
      "name": "Hypercubic",
      "url": "https://www.ycombinator.com/companies/hypercubic",
-      "cat": "C05",
+      "cat": "C04",
      "year": 2025,
      "team": 3,
      "location": "San Francisco",
@@ -10401,11 +10401,12 @@
      "stage": "目录公开，商业成熟度待核验",
      "sourceType": "YC官方页面中的公司自述",
      "evidenceIds": [
-        "YC-E-30879"
+        "YC-E-30879",
+        "TC-BF25-HYPER"
      ],
      "revenueStatus": "unknown",
      "profileOnly": true,
-      "classificationBasis": "主营简介的规则初分",
+      "classificationBasis": "人工复核YC当前主营：大型机与COBOL开发、维护和现代化工具，按开发者交付任务归C04；目录其他业务与财务字段尚未逐项核验。",
      "directoryTags": []
    },
    {
@@ -45778,6 +45779,94 @@
      "kind": "公司官方历史条款",
      "summary": "官方条款页面日期2023-10-16，服务合同方列Generated Media, Inc.，业务为生成图像许可。",
      "limitation": "单看同名条款不唯一识别主体，需与注册号7627361的隐私政策及Brila政策合看；不能把图像许可收入当新建站产品ARR。"
+    },
+    {
+      "id": "TC-ARCHIVE-2025",
+      "companyId": null,
+      "title": "TechCrunch 2025普通年度档案入口",
+      "url": "https://techcrunch.com/2025/",
+      "publishedAt": null,
+      "accessedAt": "2026-10-30",
+      "kind": "媒体档案索引",
+      "summary": "年度档案首屏可读并提供Next链接，可从普通页面继续回查2025报道。",
+      "limitation": "本次只核入口和可见首屏，不代表已穷尽分页；综合档案包含非AI内容，新闻日期不当成立年。"
+    },
+    {
+      "id": "TC-ARCHIVE-AI2",
+      "companyId": null,
+      "title": "TechCrunch AI栏目第二页读取范围",
+      "url": "https://techcrunch.com/category/artificial-intelligence/page/2/",
+      "publishedAt": null,
+      "accessedAt": "2026-10-30",
+      "kind": "媒体档案索引",
+      "summary": "普通AI栏目第二页可读，页面存在跨公司、产品和行业报道。",
+      "limitation": "只证明本次页面可读；动态相对时间和当前分页位置不是稳定的两年候选池，不以栏目名保证每条均为新AI公司。"
+    },
+    {
+      "id": "TC-BF25",
+      "companyId": null,
+      "title": "2025企业竞赛名单的定向发现入口",
+      "url": "https://techcrunch.com/2025/12/29/the-32-top-enterprise-tech-startups-from-disrupt-startup-battlefield/",
+      "publishedAt": "2025-12-29",
+      "accessedAt": "2026-10-30",
+      "kind": "媒体候选发现",
+      "summary": "从2025企业竞赛精选名单选择五条线索，逐项另查原始资料。",
+      "limitation": "仅读取并核查五项，不声称32项、全部200参赛者或两年档案完成；参赛年不等成立年，不复用文章宣传评价为独立能力判断。"
+    },
+    {
+      "id": "TC-BF25-HYPER",
+      "companyId": "YC30879",
+      "title": "Hypercubic当前YC主营与年份复核",
+      "url": "https://www.ycombinator.com/companies/hypercubic",
+      "publishedAt": null,
+      "accessedAt": "2026-10-30",
+      "kind": "YC官方公司资料",
+      "summary": "页面列Founded 2025，描述面向大型机与COBOL的维护现代化及Hopper开发环境。",
+      "limitation": "同一公司复用YC30879；公开演示、客户行业或目标企业不等付费客户，未测试性能或核收入。"
+    },
+    {
+      "id": "TC-BF25-BILLOW",
+      "companyId": null,
+      "title": "Billow AI Labs的YC资料与成立年缺项",
+      "url": "https://www.ycombinator.com/companies/billow-ai-labs",
+      "publishedAt": null,
+      "accessedAt": "2026-10-30",
+      "kind": "YC官方公司资料",
+      "summary": "页面链接thebillow.ai，列Summer 2026批次，介绍会计和财务流程服务，本次可读正文无Founded字段。",
+      "limitation": "批次不等成立年；与2025竞赛发现线索的时序和实体需补证，不仅按相同域名和批次新增2026公司。创始人旧平台用户、交易量和客户企业价值不当Billow收入。"
+    },
+    {
+      "id": "TC-BF25-DOBS",
+      "companyId": null,
+      "title": "Dobs AI公司主页的范围外年份",
+      "url": "https://www.linkedin.com/company/dobs-ai",
+      "publishedAt": null,
+      "accessedAt": "2026-10-30",
+      "kind": "公司公开主页自述",
+      "summary": "公司主页链接dobs.ai并列Founded 2024，介绍财务尽调与盈利质量分析。",
+      "limitation": "2024为公司自述，精确注册未核；准确率及合规宣传不是本轮独立测试，客户数据分析金额不作其自身收入。"
+    },
+    {
+      "id": "TC-BF25-CASHEW",
+      "companyId": null,
+      "title": "Cashew Research官方业务与同名边界",
+      "url": "https://www.cashewresearch.com/about",
+      "publishedAt": null,
+      "accessedAt": "2026-10-30",
+      "kind": "公司官方材料",
+      "summary": "公开公司介绍页可读，关联市场研究业务与cashewresearch.com。",
+      "limitation": "成立年及运营主体仍待可靠材料补证；不是YC的Cashew Labs，不能套用后者2025年份、团队或技术。"
+    },
+    {
+      "id": "TC-BF25-ZINNIA",
+      "companyId": null,
+      "title": "Zinnia发现域名的当前访问缺项",
+      "url": "https://www.getzinnia.ai/",
+      "publishedAt": null,
+      "accessedAt": "2026-10-30",
+      "kind": "公司官网访问核查",
+      "summary": "从发现链接访问getzinnia.ai，本次重定向getzinnia.com后无法读取正文。",
+      "limitation": "不以搜索摘录或其他同名Zinnia公司填补成立年、团队或收入；当前无法核实该域名的主体连续性。"
    }
  ],
  "reviews": [
@@ -47143,6 +47232,46 @@
      ],
      "handling": "归范围外既有主体；保留精确注册日未知，不因产品发布新增公司，不将图像业务收入或团队规模转作Brila建站业务。",
      "scope": "excluded-candidate"
+    },
+    {
+      "id": "TC-R-ARCHIVE-SCOPE",
+      "type": "档案入口与读取边界",
+      "company": "资料来源整体",
+      "summary": "2025普通年度档案与AI第二页可读，说明RSS/API受限不代表所有普通档案页面不可读。",
+      "evidenceIds": [
+        "TC-ARCHIVE-2025",
+        "TC-ARCHIVE-AI2",
+        "TC-BF25"
+      ],
+      "handling": "补2025定向候选台账，继续保留未读分页与未核名单；不把入口可读、竞赛选样或新闻时点写成两年新公司全集。",
+      "scope": "dataset"
+    },
+    {
+      "id": "HYPER-R-CLASSIFICATION",
+      "type": "主营人工分类复核",
+      "company": "Hypercubic",
+      "companyIds": [
+        "YC30879"
+      ],
+      "summary": "当前YC材料显示大型机/COBOL开发维护与现代化，归开发者工具。",
+      "evidenceIds": [
+        "YC-E-30879",
+        "TC-BF25-HYPER"
+      ],
+      "handling": "复用既有公司记录，不重复新增；仍为目录基础记录，商业模式、定价及财务继续留缺，不把演示当独立测试。",
+      "scope": "companies"
+    },
+    {
+      "id": "DOBS-R-COHORT",
+      "type": "参赛与成立年边界",
+      "company": "Dobs AI（既有公司候选）",
+      "summary": "2025参赛发现线索由公司主页2024自述年份补证为范围外。",
+      "evidenceIds": [
+        "TC-BF25",
+        "TC-BF25-DOBS"
+      ],
+      "handling": "不新增公司或收入观测，不将客户盈利分析结果、准确率或合规宣传转为公司收入或独立技术验证。",
+      "scope": "excluded-candidate"
    }
  ],
  "coverage": {
@@ -47886,6 +48015,87 @@
          "companyId": "AI079"
        }
      ]
+    },
+    {
+      "id": "TC-2025-battlefield-targeted-1",
+      "sourceId": "S02",
+      "url": "https://techcrunch.com/2025/12/29/the-32-top-enterprise-tech-startups-from-disrupt-startup-battlefield/",
+      "accessedAt": "2026-10-30",
+      "scope": "从2025-12-29企业竞赛名单中选取五条线索，按另列原始资料补证；不是该文32项全表、全部参赛者或两年档案。名录、参赛和报道时点均不直接作成立年。",
+      "complete": false,
+      "records": [
+        {
+          "id": "TC25-BF-01",
+          "name": "Billow AI Labs",
+          "url": "https://www.ycombinator.com/companies/billow-ai-labs",
+          "categoryId": "C07",
+          "kind": "company",
+          "profileStatus": "readable",
+          "status": "pending-year",
+          "evidenceIds": [
+            "TC-BF25",
+            "TC-BF25-BILLOW"
+          ],
+          "note": "YC仅列S2026批次且无可读Founded字段；与2025发现线索的实体/时序待核。旧创业业绩、客户企业价值和交易量不作新公司收入。"
+        },
+        {
+          "id": "TC25-BF-02",
+          "name": "Cashew Research",
+          "url": "https://www.cashewresearch.com/about",
+          "categoryId": "C06",
+          "kind": "company",
+          "profileStatus": "readable",
+          "status": "pending-year",
+          "evidenceIds": [
+            "TC-BF25",
+            "TC-BF25-CASHEW"
+          ],
+          "note": "成立年与主体未补齐；不与Cashew Labs混用，也不从参赛或新闻年推成立年。"
+        },
+        {
+          "id": "TC25-BF-03",
+          "name": "Dobs AI",
+          "url": "https://www.linkedin.com/company/dobs-ai",
+          "categoryId": "C07",
+          "kind": "company",
+          "profileStatus": "readable",
+          "status": "outside-cohort",
+          "evidenceIds": [
+            "TC-BF25",
+            "TC-BF25-DOBS"
+          ],
+          "note": "公司主页自述2024成立，按范围外保留，不凭2025参赛新增公司；精确注册与财务未核。"
+        },
+        {
+          "id": "TC25-BF-04",
+          "name": "Hypercubic",
+          "url": "https://www.ycombinator.com/companies/hypercubic",
+          "categoryId": "C04",
+          "kind": "company",
+          "profileStatus": "readable",
+          "status": "linked",
+          "evidenceIds": [
+            "TC-BF25",
+            "TC-BF25-HYPER"
+          ],
+          "note": "复用既有YC30879，2025年份由YC自述；当前主营复核为大型机开发维护工具，收入未知。",
+          "companyId": "YC30879"
+        },
+        {
+          "id": "TC25-BF-05",
+          "name": "Zinnia（getzinnia.ai线索）",
+          "url": "https://www.getzinnia.ai/",
+          "categoryId": "C06",
+          "kind": "company",
+          "profileStatus": "unavailable",
+          "status": "pending-year",
+          "evidenceIds": [
+            "TC-BF25",
+            "TC-BF25-ZINNIA"
+          ],
+          "note": "官方域名重定向后的正文不可读，成立年和主体连续性留缺，不套用其他同名公司资料。"
+        }
+      ]
    }
  ],
  "directoryRechecks": [
@@ -1,5 +1,5 @@
 {
-  "version": "2026-10-30-18",
+  "version": "2026-10-30-19",
  "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。",
@@ -28,8 +28,8 @@
        "techcrunch.com"
      ],
      "initialAccess": "AI栏目可读取",
-      "boundary": "栏目首页选择的3条公司线索已追溯原始资料：Wabi据公开名录2025年份、Thrive Holdings据公司主页2025自述关联公司，DreamHQ成立年仍待核。Thrive控股改造与收购/有机收入边界单列；非首页全部条目或两年档案，公开分类接口与RSS普通HTTP请求返回403。",
-      "nextCheck": "补可公开取得的2025、2026档案及剩余分页；核DreamHQ成立依据与Wabi、Thrive Holdings精确注册、团队及收入边界。",
+      "boundary": "栏目首页选择的3条公司线索已追溯原始资料：Wabi据公开名录2025年份、Thrive Holdings据公司主页2025自述关联公司，DreamHQ成立年仍待核。Thrive控股改造与收购/有机收入边界单列；非首页全部条目或两年档案，公开分类接口与RSS普通HTTP请求返回403。 已补读2025年度档案入口、AI第二页，并从2025企业竞赛报道定向核5条：复用1条、范围外1条、待核3条；不是完整32项名单或两年全量。",
+      "nextCheck": "沿已验证可读的普通2025档案及AI分页继续补台账；记录未读分页，补DreamHQ、Billow、Cashew Research和Zinnia成立与主体依据，区分竞赛年、批次、产品和公司成立年。",
      "limitation": "需追溯原始披露，部分文章受访问限制",
      "accessRecheckedAt": "2026-10-30"
    },
@@ -1,5 +1,5 @@
 {
-  "sourceVersion": "2026-10-30-thrive-brila-1",
+  "sourceVersion": "2026-10-30-archives-1",
  "asOf": "2026-10-30",
  "metric": "platform-mrr-usd",
  "sampleUnit": "product-or-business",