tihao-ai/claude-code-tihao-sourcing/scripts/homepage-evidence-readiness-audit.js

#!/usr/bin/env node
const fs = require('fs');
const path = require('path');

const ROOT = path.resolve(__dirname, '..');
const OUTPUTS = path.join(ROOT, 'outputs');

function main() {
  const args = parseArgs(process.argv.slice(2));
  const input = args.input || args.result || args.evidence || '';
  const outputDir = path.resolve(args.output || process.env.TIHAO_HOMEPAGE_READINESS_OUTPUT || path.join(OUTPUTS, `homepage-evidence-readiness-${Date.now()}`));
  const records = input ? loadRecords(path.resolve(input)) : loadDefaultRecords();
  const summary = buildHomepageEvidenceReadiness({
    inputPath: input ? path.resolve(input) : '',
    records
  });

  fs.mkdirSync(outputDir, { recursive: true });
  const jsonPath = path.join(outputDir, 'homepage-evidence-readiness-summary.json');
  const reportPath = path.join(outputDir, 'homepage-evidence-readiness-report.md');
  const repairCsvPath = path.join(outputDir, 'homepage-evidence-repair-actions.csv');
  fs.writeFileSync(jsonPath, JSON.stringify(summary, null, 2), 'utf8');
  fs.writeFileSync(reportPath, withBom(renderReport(summary)), 'utf8');
  fs.writeFileSync(repairCsvPath, withBom(renderRepairCsv(summary.repairActions)), 'utf8');

  console.log(JSON.stringify({
    outputDir,
    json: jsonPath,
    report: reportPath,
    repairCsv: repairCsvPath,
    ready: summary.ready,
    passed: summary.passed,
    proofLevel: summary.proofLevel,
    candidateCount: summary.counts.candidates,
    providerEvidenceCount: summary.counts.providerEvidenceCreators,
    failureCount: summary.failureCount,
    repairActionCount: summary.repairActions.length
  }, null, 2));

  if (args.strict && !summary.ready) process.exitCode = 2;
}

function buildHomepageEvidenceReadiness({ inputPath, records }) {
  const candidates = records.map((record, index) => normalizeCandidate(record, index + 1));
  const providerEvidenceCreators = candidates.filter(item => item.hasProviderEvidence);
  const creatorsWithPosts = candidates.filter(item => item.recentPostCount > 0);
  const creatorsWithEnoughRecentPosts = candidates.filter(item => item.recentPostCount >= 10);
  const creatorsWithTitleOrText = candidates.filter(item => item.postTitleOrTextCount > 0 || item.reviewNoteCount > 0);
  const creatorsWithCover = candidates.filter(item => item.coverCount > 0 || item.visualQualityScore > 0);
  const creatorsWithPublishTime = candidates.filter(item => item.publishTimeCount > 0);
  const creatorsWithInteractions = candidates.filter(item => item.interactionCount > 0);
  const creatorsWithRiskReview = candidates.filter(item => item.riskSignalCount > 0 || item.reviewNoteCount > 0);

  const acceptance = {
    hasCandidates: candidates.length > 0,
    hasProviderOrPathEvidence: providerEvidenceCreators.length > 0,
    hasRecentPosts: creatorsWithPosts.length > 0,
    recentWindowUsable: providerEvidenceCreators.length > 0 && providerEvidenceCreators.every(item => item.recentContentWindow >= 10 || item.recentPostCount >= 10),
    hasTitleOrText: creatorsWithTitleOrText.length > 0,
    hasCoverEvidence: creatorsWithCover.length > 0,
    hasPublishTime: creatorsWithPublishTime.length > 0,
    hasInteractionEvidence: creatorsWithInteractions.length > 0,
    preservesRiskReview: creatorsWithRiskReview.length > 0,
    noSecrets: !containsSecret(JSON.stringify(records))
  };
  const ready = Object.values(acceptance).every(Boolean);
  const issues = buildIssues({ acceptance, candidates, providerEvidenceCreators });
  const repairActions = buildRepairActions({ issues, inputPath });

  return {
    generatedAt: new Date().toISOString(),
    inputPath,
    ready,
    passed: true,
    complete: ready,
    directCustomerProof: false,
    proofLevel: ready ? 'smoke_or_local' : 'not_business_proof',
    counts: {
      candidates: candidates.length,
      providerEvidenceCreators: providerEvidenceCreators.length,
      creatorsWithPosts: creatorsWithPosts.length,
      creatorsWithEnoughRecentPosts: creatorsWithEnoughRecentPosts.length,
      creatorsWithTitleOrText: creatorsWithTitleOrText.length,
      creatorsWithCover: creatorsWithCover.length,
      creatorsWithPublishTime: creatorsWithPublishTime.length,
      creatorsWithInteractions: creatorsWithInteractions.length,
      creatorsWithRiskReview: creatorsWithRiskReview.length
    },
    acceptance,
    failureCount: issues.length,
    issues,
    repairActions,
    sample: candidates.slice(0, 10).map(item => ({
      rowNumber: item.rowNumber,
      platform: item.platform,
      displayName: item.displayName,
      profileUrl: item.profileUrl,
      evidenceSource: item.evidenceSource,
      recentPostCount: item.recentPostCount,
      recentContentWindow: item.recentContentWindow,
      coverCount: item.coverCount,
      publishTimeCount: item.publishTimeCount,
      interactionCount: item.interactionCount,
      riskSignalCount: item.riskSignalCount
    }))
  };
}

function loadDefaultRecords() {
  const candidates = [
    path.join(OUTPUTS, 'software-client-latest', 'software-client-list.final.json'),
    path.join(OUTPUTS, 'software-client-latest', 'tihao-sourcing-result.json'),
    path.join(OUTPUTS, 'claude-code-tihao-sample', 'tihao-sourcing-result.json')
  ];
  for (const file of candidates) {
    if (fs.existsSync(file)) return loadRecords(file);
  }
  return [];
}

function loadRecords(file) {
  if (!fs.existsSync(file)) return [];
  if (/\.csv$/i.test(file)) return readCsv(file).body;
  const json = readJson(file);
  if (!json) return [];
  return normalizeJsonRecords(json);
}

function normalizeJsonRecords(json) {
  if (Array.isArray(json)) return json;
  const pools = [
    json.candidates,
    json.creators,
    json.records,
    json.homepageEvidence,
    json.evidence,
    json.data?.candidates,
    json.data?.creators,
    json.data?.records,
    json.data?.homepageEvidence,
    json.result?.candidates,
    json.summary?.candidates
  ];
  return pools.find(Array.isArray) || [];
}

function normalizeCandidate(record, rowNumber) {
  const homepageEvidence = record.homepageEvidence || record.homepage || record.evidence || {};
  const posts = normalizePosts(record.recentPosts || record.posts || homepageEvidence.recentPosts || homepageEvidence.posts || []);
  const reviewNotes = normalizeList(homepageEvidence.reviewNotes || record.recentEvidence || record.reviewNotes);
  const riskHits = normalizeList(homepageEvidence.riskHits || homepageEvidence.visualRiskSignals || record.homepageQualityRisks || record.riskSignals);
  const evidenceSource = value(homepageEvidence.source || record.homepageEvidenceSource || record.source);
  const recentContentWindow = number(homepageEvidence.recentContentWindow || homepageEvidence.recentPostCount || record.recentContentWindow || posts.length);
  const visualQualityScore = number(homepageEvidence.visualQualityScore || record.visualQualityScore);
  const coverCount = posts.filter(hasCover).length + number(homepageEvidence.coverCount || record.coverCount);
  const postTitleOrTextCount = posts.filter(post => value(post.title || post.desc || post.description || post.text || post.content || post.summary).length > 0).length;
  const publishTimeCount = posts.filter(post => value(post.publishTime || post.publishedAt || post.createdAt || post.time).length > 0).length;
  const interactionCount = posts.filter(hasInteraction).length;
  const hasProviderEvidence = ['provider', 'path', 'api', 'social-analysis', 'voc-e-commerce', 'voc-social'].includes(evidenceSource.toLowerCase()) ||
    Boolean(record.recentPosts || record.posts || homepageEvidence.recentPosts || homepageEvidence.posts) ||
    homepageEvidence.confidence === 'provider';

  return {
    rowNumber,
    platform: value(record.platform || record['平台']),
    displayName: value(record.displayName || record.creatorName || record.name || record['博主名称']),
    profileUrl: value(record.profileUrl || record.homepageUrl || record.url || record['主页链接']),
    evidenceSource: evidenceSource || (hasProviderEvidence ? 'record' : 'fallback'),
    hasProviderEvidence,
    recentPostCount: posts.length,
    recentContentWindow,
    postTitleOrTextCount,
    coverCount,
    publishTimeCount,
    interactionCount,
    riskSignalCount: riskHits.length,
    reviewNoteCount: reviewNotes.length,
    visualQualityScore
  };
}

function normalizePosts(posts) {
  if (!Array.isArray(posts)) return [];
  return posts.filter(item => item && typeof item === 'object');
}

function buildIssues({ acceptance, candidates, providerEvidenceCreators }) {
  const issues = [];
  if (!acceptance.hasCandidates) issues.push(issue('missing-candidates', '没有可审计的候选博主记录。'));
  if (!acceptance.hasProviderOrPathEvidence) issues.push(issue('missing-provider-evidence', '没有 creator 具备 provider/path 级主页近期内容证据。'));
  if (!acceptance.hasRecentPosts) issues.push(issue('missing-recent-posts', '没有近期内容列表，不能判断最近 10/20 篇内容。'));
  if (!acceptance.recentWindowUsable) issues.push(issue('insufficient-recent-window', 'provider/path 证据的近期内容窗口不足 10 篇。'));
  if (!acceptance.hasTitleOrText) issues.push(issue('missing-title-or-text', '近期内容缺少标题、正文或摘要。'));
  if (!acceptance.hasCoverEvidence) issues.push(issue('missing-cover-evidence', '缺少封面或视觉质感证据。'));
  if (!acceptance.hasPublishTime) issues.push(issue('missing-publish-time', '近期内容缺少发布时间。'));
  if (!acceptance.hasInteractionEvidence) issues.push(issue('missing-interactions', '近期内容缺少互动字段。'));
  if (!acceptance.preservesRiskReview) issues.push(issue('missing-risk-review', '缺少风险信号或复核备注。'));
  if (!acceptance.noSecrets) issues.push(issue('secret-like-value', '主页证据输入中包含疑似 token 或鉴权字段。'));

  for (const candidate of providerEvidenceCreators) {
    if (candidate.recentPostCount > 0 && candidate.recentPostCount < 10) {
      issues.push(issue('creator-window-too-small', `${candidate.displayName || candidate.profileUrl || `row ${candidate.rowNumber}`} 的近期内容少于 10 篇。`, candidate.rowNumber));
    }
    if (candidate.recentPostCount > 0 && candidate.publishTimeCount === 0) {
      issues.push(issue('creator-missing-publish-time', `${candidate.displayName || candidate.profileUrl || `row ${candidate.rowNumber}`} 的近期内容没有发布时间。`, candidate.rowNumber));
    }
  }
  return issues;
}

function buildRepairActions({ issues, inputPath }) {
  return issues.map((item, index) => {
    const spec = repairSpec(item.type);
    return {
      priority: index + 1,
      owner: spec.owner,
      type: item.type,
      rowNumber: item.rowNumber || '',
      file: inputPath || 'outputs/software-client-latest/software-client-list.final.csv',
      field: spec.field,
      action: spec.action,
      acceptance: spec.acceptance
    };
  });
}

function repairSpec(type) {
  const map = {
    'missing-candidates': ['技术/AI', '候选名单', '传入 tihao-sourcing-result.json 或包含候选博主的 CSV/JSON。', '候选记录数大于 0。'],
    'missing-provider-evidence': ['技术/AI', 'homepageEvidence.source', '接入真实主页近期内容 provider 或提供 path/local JSON 证据。', '至少 1 个 creator 的 evidenceSource 为 provider/path/api。'],
    'missing-recent-posts': ['商务/投放', 'recentPosts', '补最近 10/20 篇内容列表。', 'recentPosts 条数大于 0，强推荐候选优先达到 10 篇以上。'],
    'insufficient-recent-window': ['商务/投放', 'recentPosts', '补足最近 10 篇以上主页内容。', 'provider/path 证据的 recentPostCount 或 recentContentWindow >= 10。'],
    'missing-title-or-text': ['商务/投放', 'title/text/summary', '补每篇内容的标题、正文或摘要。', '至少 1 篇近期内容有标题、正文或摘要。'],
    'missing-cover-evidence': ['商务/投放', 'coverUrl/visualQualityScore', '补封面链接或视觉质感判断。', '至少 1 篇近期内容有封面，或主页证据保留 visualQualityScore。'],
    'missing-publish-time': ['商务/投放', 'publishTime', '补近期内容发布时间。', '至少 1 篇近期内容有发布时间；provider 声明 recent_posts 时应尽量全量保留。'],
    'missing-interactions': ['商务/投放', 'likes/comments/shares', '补点赞、评论、收藏、分享等互动字段。', '至少 1 篇近期内容有互动字段。'],
    'missing-risk-review': ['技术/AI', 'riskHits/reviewNotes', '保留主页下沉、封面混乱、调性不符等风险信号。', 'summary 中 preservesRiskReview=true。'],
    'secret-like-value': ['技术/AI', 'secrets', '移除 token、Authorization、sessionToken 等敏感字段。', 'noSecrets=true。'],
    'creator-window-too-small': ['商务/投放', 'recentPosts', '补足该博主最近 10 篇主页内容。', '该 creator 的 recentPostCount >= 10。'],
    'creator-missing-publish-time': ['商务/投放', 'publishTime', '补该博主近期内容发布时间。', '该 creator 的 publishTimeCount > 0。']
  };
  const value = map[type] || ['技术/AI', 'homepageEvidence', '补齐主页证据字段。', '对应 gate 通过。'];
  return { owner: value[0], field: value[1], action: value[2], acceptance: value[3] };
}

function renderReport(summary) {
  const lines = [
    '# 主页近期内容证据就绪审计',
    '',
    `- 生成时间：${summary.generatedAt}`,
    `- 是否 ready：${summary.ready ? '是' : '否'}`,
    `- proofLevel：${summary.proofLevel}`,
    `- directCustomerProof：${summary.directCustomerProof ? 'true' : 'false'}`,
    `- failureCount：${summary.failureCount}`,
    '',
    '## 统计',
    '',
    `- 候选博主数：${summary.counts.candidates}`,
    `- provider/path 证据博主数：${summary.counts.providerEvidenceCreators}`,
    `- 有近期内容博主数：${summary.counts.creatorsWithPosts}`,
    `- 最近内容窗口 >= 10 的博主数：${summary.counts.creatorsWithEnoughRecentPosts}`,
    `- 有标题/正文/摘要证据博主数：${summary.counts.creatorsWithTitleOrText}`,
    `- 有封面/视觉证据博主数：${summary.counts.creatorsWithCover}`,
    `- 有发布时间博主数：${summary.counts.creatorsWithPublishTime}`,
    `- 有互动字段博主数：${summary.counts.creatorsWithInteractions}`,
    `- 有风险复核博主数：${summary.counts.creatorsWithRiskReview}`,
    '',
    '## 门禁',
    '',
    '| 门禁 | 状态 |',
    '| --- | --- |',
    ...Object.entries(summary.acceptance).map(([key, pass]) => `| ${key} | ${pass ? 'pass' : 'fail'} |`),
    '',
    '## 问题',
    '',
    summary.issues.length
      ? '| 类型 | 行号 | 说明 |\n| --- | --- | --- |\n' + summary.issues.map(item => `| ${item.type} | ${item.rowNumber || ''} | ${escapeCell(item.message)} |`).join('\n')
      : '- 暂无问题。',
    '',
    '## 修复清单',
    '',
    summary.repairActions.length
      ? '| 优先级 | 负责人 | 字段 | 修复动作 | 通过标准 |\n| ---: | --- | --- | --- | --- |\n' + summary.repairActions.map(item => `| ${item.priority} | ${item.owner} | ${escapeCell(item.field)} | ${escapeCell(item.action)} | ${escapeCell(item.acceptance)} |`).join('\n')
      : '- 暂无修复动作。',
    '',
    '## 边界',
    '',
    '- 本审计只判断主页近期内容证据是否足够支撑强推荐复核，不证明客户命中率提升。',
    '- provider/path 证据不足时，fallback 只能作为轻量判断，不能单独支撑强推荐。',
    '- sample、smoke、接口 200 或 provider fallback 不能当成真实业务效果证明。',
    '- 输出不得包含 sessionToken、Authorization、模型 token 或 npm token。'
  ];
  return lines.join('\n');
}

function renderRepairCsv(actions) {
  const header = ['优先级', '负责人', '类型', '行号', '文件', '字段', '修复动作', '通过标准'];
  return [
    header.join(','),
    ...actions.map(item => [
      item.priority,
      item.owner,
      item.type,
      item.rowNumber,
      item.file,
      item.field,
      item.action,
      item.acceptance
    ].map(csvCell).join(','))
  ].join('\n');
}

function readJson(file) {
  try {
    return JSON.parse(fs.readFileSync(file, 'utf8').replace(/^\uFEFF/, ''));
  } catch {
    return null;
  }
}

function readCsv(file) {
  const text = fs.readFileSync(file, 'utf8').replace(/^\uFEFF/, '');
  const lines = text.split(/\r?\n/).filter(Boolean).map(parseCsvLine);
  const header = lines[0] || [];
  const body = lines.slice(1).map(values => Object.fromEntries(header.map((key, index) => [key, values[index] || ''])));
  return { header, body };
}

function parseCsvLine(line) {
  const cells = [];
  let current = '';
  let quoted = false;
  for (let index = 0; index < line.length; index += 1) {
    const char = line[index];
    if (char === '"' && line[index + 1] === '"') {
      current += '"';
      index += 1;
    } else if (char === '"') {
      quoted = !quoted;
    } else if (char === ',' && !quoted) {
      cells.push(current);
      current = '';
    } else {
      current += char;
    }
  }
  cells.push(current);
  return cells;
}

function normalizeList(value) {
  if (Array.isArray(value)) return value.filter(Boolean).map(String);
  if (!value) return [];
  return String(value).split(/[|,，、\n]/).map(item => item.trim()).filter(Boolean);
}

function hasCover(post) {
  return value(post.coverUrl || post.cover || post.imageUrl || post.thumbnail || post.noteCover).length > 0;
}

function hasInteraction(post) {
  return ['likeCount', 'likes', 'commentCount', 'comments', 'collectCount', 'favorites', 'shareCount', 'shares', 'viewCount', 'views'].some(key => number(post[key]) > 0 || value(post[key]).length > 0);
}

function containsSecret(text) {
  return /(sessionToken|Authorization|Bearer\s+|sk-[A-Za-z0-9_-]{12,}|npm_[A-Za-z0-9_-]{12,}|r:[A-Za-z0-9]{20,})/i.test(String(text || ''));
}

function issue(type, message, rowNumber = null) {
  return { type, message, rowNumber };
}

function value(input) {
  return String(input ?? '').trim();
}

function number(input) {
  const parsed = Number(input);
  return Number.isFinite(parsed) ? parsed : 0;
}

function escapeCell(input) {
  return value(input).replace(/\|/g, '/').replace(/\n/g, ' ');
}

function csvCell(input) {
  const text = value(input);
  return /[",\n]/.test(text) ? `"${text.replace(/"/g, '""')}"` : text;
}

function withBom(text) {
  return `\uFEFF${text}`;
}

function parseArgs(argv) {
  const result = {};
  for (let index = 0; index < argv.length; index += 1) {
    const arg = argv[index];
    if (!arg.startsWith('--')) continue;
    const key = arg.slice(2).replace(/-([a-z])/g, (_, char) => char.toUpperCase());
    const next = argv[index + 1];
    if (!next || next.startsWith('--')) result[key] = true;
    else {
      result[key] = next;
      index += 1;
    }
  }
  return result;
}

if (require.main === module) {
  try {
    main();
  } catch (error) {
    console.error(error && error.stack ? error.stack : String(error));
    process.exit(1);
  }
}

module.exports = {
  buildHomepageEvidenceReadiness,
  normalizeCandidate
};