memind/scripts/news-item-generate-page.mjs

#!/usr/bin/env node
/**
 * 阶段一：Agent 深度分析单条新闻并生成落地页。只生成，不推送。
 *
 * stdin  { userId, newsItem, runId? }
 * stdout { ok, pageSlug, pageUrl, relativePath, publishDir, html, check, executor }
 *
 * 落地页生成后即可点链接查看；后续的预览、编辑、推送都不再需要跑 Agent。
 */
import fs from 'node:fs';
import path from 'node:path';
import process from 'node:process';
import { executeScheduledTask } from '../scheduled-task-executor.mjs';
import {
  buildPublicUrl,
  PUBLISH_ROOT_DIR,
  resolvePublicBaseUrl,
} from '../user-publish.mjs';
import {
  buildNewsItemGenerationTaskSpec,
  validateNewsItemHtml,
} from '../news-item-templates.mjs';
import {
  bootstrapAgentContext,
  loadPushEnvironment,
  readStdinPayload,
  runScript,
} from './news-item-wechat-context.mjs';

function slugify(value) {
  return String(value ?? '')
    .trim()
    .toLowerCase()
    .replace(/[^\p{L}\p{N}]+/gu, '-')
    .replace(/^-+|-+$/g, '')
    .slice(0, 48) || 'news-item';
}

function buildGenerationTask(userId, newsItem, env, now = Date.now()) {
  const date = new Date(now);
  const mmdd = `${String(date.getMonth() + 1).padStart(2, '0')}${String(date.getDate()).padStart(2, '0')}`;
  const slug = `news-item-${slugify(newsItem.canonicalTitle || newsItem.title)}-${mmdd}`;
  const { style, taskSpec } = buildNewsItemGenerationTaskSpec(newsItem, env);
  return {
    id: `news-item-${slug}`,
    userId,
    title: `新闻深度稿：${newsItem.canonicalTitle || newsItem.title}`,
    taskSpec: [
      taskSpec,
      '',
      `输出文件名必须是 public/${slug}.html（可覆盖同名旧文件）。`,
      `data-style 必须设为 ${style}。`,
    ].join('\n'),
    recurrence: 'once',
    timezone: 'Asia/Shanghai',
    notifyChannel: 'web',
    expectedSlug: slug,
    expectedRelativePath: `public/${slug}.html`,
    layoutStyle: style,
    isExpectedPageReady: (html) => validateNewsItemHtml(html, { style }).ok,
  };
}

function resolveTimeoutMs(env) {
  return Math.max(
    5 * 60_000,
    Number(env.MEMIND_NEWS_ITEM_GENERATE_TIMEOUT_MS ?? 20 * 60_000) || 20 * 60_000,
  );
}

async function main() {
  const env = loadPushEnvironment(process.env);
  const payload = readStdinPayload();
  const userId = String(payload.userId ?? '').trim();
  const newsItem = payload.newsItem ?? {};
  if (!userId) throw new Error('缺少 userId');

  const services = await bootstrapAgentContext(env);
  try {
    const task = buildGenerationTask(userId, newsItem, env);
    const execution = await executeScheduledTask(task, {
      userAuth: services.userAuth,
      tkmindProxy: services.tkmindProxy,
      agentRunGateway: services.agentRunGateway,
      cursorExecutorPolicyService: services.cursorExecutorPolicyService,
      sessionSnapshotService: services.sessionSnapshotService,
      pool: services.pool,
      h5Root: services.h5Root,
      timeoutMs: resolveTimeoutMs(env),
      logger: console,
    });

    const publishDir = path.join(services.h5Root, PUBLISH_ROOT_DIR, userId);
    const relativePath = execution.readyPaths?.[0]
      ?? (fs.existsSync(path.join(publishDir, task.expectedRelativePath))
        ? task.expectedRelativePath
        : null);
    if (!relativePath) {
      throw new Error('Agent 未生成 public/*.html 页面');
    }

    const localPath = path.join(publishDir, relativePath);
    if (!fs.existsSync(localPath)) throw new Error(`生成页面不存在：${relativePath}`);
    const html = fs.readFileSync(localPath, 'utf8');
    const check = validateNewsItemHtml(html, { style: task.layoutStyle });
    if (!check.ok) {
      throw Object.assign(
        new Error(`生成页面不达标：${check.issues.join('；')}`),
        { code: 'news_item_page_substandard' },
      );
    }

    return {
      pageSlug: path.basename(relativePath, '.html'),
      pageUrl: buildPublicUrl(resolvePublicBaseUrl(env), userId, relativePath),
      relativePath,
      publishDir,
      html,
      check,
      layoutStyle: task.layoutStyle,
      executor: execution.executor ?? null,
    };
  } finally {
    await services.close();
  }
}

runScript(main);