Refactor: scripts own file I/O, agent only produces summaries

Root cause: harvest.js fetched content but wrote NO files — it dumped a
giant manifest (with full body) to stdout and left all document I/O to
the agent. This caused silent save failures (agent hand-wrote fragile
Node scripts) and token waste (full body in context → 13 python re-parses).

Architecture change — scripts do all deterministic I/O, full body never
enters the agent context:

  Archive:  harvest.js <src> <n>  → writes 原文.md + registry + index
                                     (sentinel placeholder for 中文标题)
                                     + compact manifest (bodyPreview only)
            agent writes .summaries.json (中文标题 + 四段摘要 + 分类)
            finalize.js <dayDir>    → writes 摘要.md, patches sentinels,
                                     updates registry/index, cleans up

  Preview:  harvest.js <src> <n> --preview [--full]  → read-only, no writes

Changes:
- harvest.js: wire up generateOriginalDoc/updateIndex/appendRegistry
  (were defined but never called); compact manifest (no full body);
  sentinel placeholder links; --preview/--full read-only mode; real
  asset count (was "articles with images"); empty-alt guard (no bare **);
  manifest merge across runs (fixes dangling-sentinel data loss);
  --help; module exports for finalize.js
- finalize.js (new): consumes .manifest.json + .summaries.json, writes
  summary docs, patches sentinel links, updates registry/index, cleans
  up temp files; missing summary → warn + skip (non-fatal); --help
- fetcher-direct.js: author extraction via JSON-LD → <meta name=author>
  → rel=author (was matching arbitrary JSON-LD names → empty authors);
  body noise filter (share buttons, boilerplate); image tracking-pixel
  + duplicate filtering
- fetcher-browser.js: CDP port 9223 → 9222 (match docs/chromium launch)
- SKILL.md: 3-step archive workflow, --preview flow, First-Run Setup
  onboarding, guardrails, .summaries.json contract, sentinel docs
- README.md: architecture diagram, division of labor, task flows, limits

README.md

@@ -21,21 +21,30 @@
 ## 工作原理
 
 ```
-config.json ──▶ harvest.js ──▶ fetcher-*.js ──▶ 文章数据(JSON)
+config.json ──▶ harvest.js ──▶ fetcher-*.js ──▶ 文章数据
   (数据源)       (主调度器)      (direct/browser)     │
        ▲              │                               ▼
-       │         dedup 去重                         Agent(LLM)
-       │         (registry)                    翻译/摘要/分类
-                                                   │
-                                                   ▼
-                                              Obsidian vault
+       │         dedup 去重                  写盘: 原文.md + registry + index
+       │         (registry)                  (中文标题用占位哨兵, 全文已落盘)
+       │                                            │
+       │                              stdout: 紧凑 manifest (无全文, 仅预览)
+       │                                            │
+       │                                            ▼
+       │                                       Agent(LLM)
+       │                                  读 manifest → 写 .summaries.json
+       │                                  (中文标题 + 四段摘要 + 分类)
+       │                                            │
+       └────────────────────────────────────────────┘
+                          finalize.js
+            写摘要.md / 修补原文哨兵链接 / 更新 registry+index
 ```
 
 **脚本与 Agent 分工**：
-- **脚本**（`scripts/`）：确定性工作 —— HTTP 请求、HTML 解析、图片下载、去重、文件写入
-- **Agent**（LLM）：智能工作 —— 中英翻译、生成中文摘要、文章分类
+- **脚本**（`scripts/`）：确定性工作 —— HTTP 请求、HTML 解析、图片下载、去重、**写原文.md / registry / index**（中文标题先用哨兵占位，全文直接落盘）
+- **Agent**（LLM）：智能工作 —— 读紧凑 manifest（无全文，仅 `bodyPreview` ~4000 字），生成中文标题 + 四段摘要 + 分类，写入 `.summaries.json`
+- **finalize.js**：读 `.summaries.json`，写中文摘要文档，把原文里的哨兵替换成真实标题，更新 registry/index，清理临时文件
 
-`harvest.js` 完成采集后输出 JSON manifest，提示 Agent 接管翻译与文档生成。
+> 全文 body 由脚本直接写盘，**永不进入 Agent 上下文**——这是降低 token 消耗的关键。
 
 ---
 
@@ -52,7 +61,8 @@ news-harvester/
 │   ├── config-template.json          # config.json 的模板（首次运行自动复制）
 │   └── source-management.md          # 数据源管理详细说明
 ├── scripts/
-│   ├── harvest.js                    # 主入口：node harvest.js <source_id> <count>
+│   ├── harvest.js                    # 主入口：node harvest.js <source_id> <count> [--preview] [--full]
+│   ├── finalize.js                   # 补全中文摘要：node finalize.js <dayDir>
 │   ├── source-manage.js              # 数据源增删改查
 │   ├── fetcher-direct.js             # 纯 HTTP 抓取（开放站点）
 │   └── fetcher-browser.js            # Chromium CDP 抓取（JS 渲染站点）
@@ -129,26 +139,38 @@ node scripts/source-manage.js add \
 
 或让 Agent 自然语言完成：「添加 The Guardian 作为新闻源」。
 
-#### 任务 2：采集最新文章，返回 JSON（不归档）
+#### 任务 2：采集最新文章，返回数据（不归档）
 
-适合只取数据、不落盘的场景。直接调用 fetcher：
+适合只取数据、不落盘的场景。用 `--preview` 一次调用取齐文章清单（标题/日期/作者/url/字数/图片），不写任何文件：
 
 ```bash
-# 抓取首页文章链接
-node -e "require('./scripts/fetcher-direct').fetchHomeLinks('https://theconversation.com/us','theconversation\\\\.com/[\\\\w-]+-\\\\d+$').then(l=>console.log(JSON.stringify(l,null,2)))"
+# 预览最新 3 篇（含正文预览 ~4000 字，无全文）
+node scripts/harvest.js the-conversation 3 --preview
 
-# 抓取单篇文章全文
-node scripts/fetcher-direct.js https://theconversation.com/some-article-12345
+# 预览并附带完整正文（用户明确要全文时）
+node scripts/harvest.js the-conversation 3 --preview --full
 ```
 
-#### 任务 3：完整采集并归档到 Obsidian
+#### 任务 3：完整采集并归档到 Obsidian（三步）
 
 ```bash
+# Step 1: 采集 + 写原文/registry/index（脚本包揽确定性 I/O）
 node scripts/harvest.js the-conversation 5
 ```
 
-`harvest.js` 会：抓取链接 → 去重 → 逐篇抓取全文 → 下载图片 → 输出 JSON manifest。
-然后 Agent 接管：生成中文摘要 → 写入 Obsidian 文档 → 更新 index.md / registry.json。
+脚本输出**紧凑 manifest**（无全文，仅 `bodyPreview`）。Agent 读 manifest 后：
+
+```bash
+# Step 2: Agent 生成中文摘要, 写入 <dayDir>/.summaries.json
+#   [{"url":"...","titleZh":"中文标题","category":"...","tags":[...],"summaryZh":"## 事件概述\n..."}]
+```
+
+```bash
+# Step 3: 补全中文摘要文档 + 修补原文链接 + 更新 registry/index
+node scripts/finalize.js "<dayDir>"
+```
+
+> Agent **不要手写文件 I/O**——原文、registry、index 由 `harvest.js` 写；摘要文档由 `finalize.js` 写。Agent 只产出 `.summaries.json`。
 
 ### 无头模式调用示例
 
@@ -214,13 +236,14 @@ YYYYMMDD/
 ## 限制与注意
 
 - **正文提取依赖正则**：网站改版可能导致提取失败（正文 < 100 字会被跳过）
-- **中文标题需 Agent 翻译**：`harvest.js` 只生成占位标题 `[需翻译] ...`，真正翻译由 Agent 完成
+- **中文标题/摘要由 Agent 生成**：`harvest.js` 在原文/registry/index 中用哨兵 `__SUMMARY_<slug>__` 占位，`finalize.js` 读取 Agent 写的 `.summaries.json` 后替换为真实标题并生成摘要文档
 - **browser 方法依赖 Chromium CDP**（端口 9222），需预先启动：
   ```bash
   chromium --no-sandbox --remote-debugging-port=9222 \
     --user-data-dir=/tmp/chromium-harvest --disable-gpu "about:blank" &>/dev/null &
   ```
 - **图片下载无重试**：失败仅跳过
+- **Token 卫生**：归档任务为「harvest → 写 .summaries.json → finalize」三步；不要为同一源/日重跑 harvest 仅为"多取几篇"（registry 已去重，二次跑只取真正的新文章并安全合并进待处理 manifest）；不要用 Read 反复读原文大文件——manifest 里的 `bodyPreview`（~4000 字）足够写摘要
 
 ---
 

SKILL.md

@@ -11,21 +11,77 @@ Fetch latest articles from configured news sources, archive them to Obsidian vau
 
 | Task | Command |
 |------|---------|
-| Run harvest | `node harvest.js <source_id> <count>` |
+| Archive latest articles | `node harvest.js <source_id> <count>` |
+| Preview latest (read-only) | `node harvest.js <source_id> <count> --preview` |
+| Preview with full body | `node harvest.js <source_id> <count> --preview --full` |
+| Finalize Chinese summaries | `node finalize.js <dayDir>` |
 | Add source | `node source-manage.js add` |
 | List sources | `node source-manage.js list` |
 | Edit source | `node source-manage.js edit <id>` |
 
-> Scripts are invoked from the skill root directory.
+> Scripts are invoked from the skill root directory (`scripts/` is relative to this skill's root).
 
-Scripts live in: `scripts/` (relative to this skill's root directory)
-Config file: `config.json` (auto-generated from `references/config-template.json` on first run — see Configuration below)
+**Division of labor:** scripts do all deterministic I/O (fetch, parse, download images, write 原文.md / registry / index). The agent does only the intelligent part — Chinese translation + 4-section summaries — and hands them back as a JSON file for `finalize.js` to stitch in. The full article body is written to disk by the script and **never enters the agent's context**.
+
+Config file: `config.json` (auto-generated from `references/config-template.json` on first run — see First-Run Setup / Configuration below)
+
+---
+
+## First-Run Setup
+
+The agent MUST proactively check configuration state at the start of **any** news-harvester task. If unconfigured, run this onboarding **before** doing anything else — do not wait for the user to ask. This is the single biggest cause of silent failures (articles written into a non-existent `/path/to/...` dir).
+
+**Detect unconfigured state** — either condition triggers onboarding:
+- `config.json` does not exist, OR
+- `vaultPath` still equals the template placeholder `/path/to/your/obsidian/vault`
+
+```bash
+node -e "const c=require('./config.json'); console.log(c.vaultPath==='/path/to/your/obsidian/vault'?'UNCONFIGURED':'OK')" 2>/dev/null || echo "UNCONFIGURED (no config.json)"
+```
+
+### Onboarding steps
+
+**1. Set the Obsidian vault path.** Ask the user for the absolute path to their Obsidian vault root (e.g. `~/obsidian/MyVault`). Verify it exists and is writable; create it if they intend a fresh folder:
+```bash
+mkdir -p "<vaultPath>" && test -w "<vaultPath>" && echo "writable"
+```
+Then write the real path into `config.json` (replace the placeholder). Never leave the placeholder in place.
+
+**2. Confirm the first test source.** The template ships `the-conversation` (method `direct`, no Chromium needed) — the ideal first test source. `node source-manage.js list` should show it enabled. If absent, add it:
+```bash
+node source-manage.js add --id the-conversation --name "The Conversation" \
+  --url https://theconversation.com/us --method direct --folder "The Conversation"
+```
+If the user wants a different first source (e.g. `reuters` with `browser` method), add that instead and note the Chromium prerequisite below applies.
+
+**3. Check prerequisites for the source's `method`.**
+- `direct` → nothing extra.
+- `browser` → Chromium must be running on port 9222:
+  ```bash
+  curl -s http://localhost:9222/json/version 2>/dev/null | python3 -c "import sys,json; d=json.load(sys.stdin); print('OK:', d.get('Browser',''))"
+  ```
+  If not running:
+  ```bash
+  chromium --no-sandbox --remote-debugging-port=9222 --user-data-dir=/tmp/chromium-harvest --disable-gpu "about:blank" &>/dev/null & sleep 5
+  ```
+
+**4. Smoke test (read-only, no vault writes).** Confirm the fetch pipeline works end-to-end before archiving anything:
+```bash
+node harvest.js the-conversation 2 --preview
+```
+If this returns article data, setup is complete. If it errors (empty list, parse failure, network), fix the source config before proceeding — do not move on to a full archive.
+
+**5. (Optional) full test archive.** Only if the user wants to verify vault writes — run Flow A once with `count 1-2` and confirm files land in `<vaultPath>/<vaultFolder>/YYYYMMDD/`. Skip otherwise; the `--preview` smoke test is enough to consider setup done.
+
+### After onboarding
+
+Once `vaultPath` is real and the smoke test passes, **skip this section** on future runs and go straight to Workflow. The check is idempotent and cheap — re-running `node source-manage.js list` is always safe.
 
 ---
 
 ## Configuration
 
-`config.json` is **not shipped** with this skill — it is machine-specific (contains the local vault path). On first run, any script auto-creates `config.json` by copying `references/config-template.json`.
+`config.json` is **not shipped** with this skill — it is machine-specific (contains the local vault path). On first run, any script auto-creates `config.json` by copying `references/config-template.json`. For the guided first-time setup flow, see **First-Run Setup** above.
 
 Before running a harvest, the agent MUST verify/edit these fields:
 
@@ -40,49 +96,117 @@ To point a source at a different vault subfolder, set `sources[].vaultFolder`. A
 
 ## Workflow
 
-When asked to harvest news, follow these steps:
+Two flows depending on intent: **archive** (fetch + save to Obsidian) or **preview** (read-only, just return article data).
 
-### 1. Load config
-Read `config.json`. If it doesn't exist, create it from the default template in `references/config-template.json`.
+### Flow A — Archive to Obsidian (3 steps)
 
-### 2. Check Chromium (for browser sources)
+#### Step 1. Run harvest (script writes originals + registry + index)
 ```bash
-curl -s http://localhost:9222/json/version 2>/dev/null | python3 -c "import sys,json; d=json.load(sys.stdin); print('OK:', d.get('Browser',''))"
+node harvest.js <source_id> <count>
 ```
-If not running, start it:
-```bash
-chromium --no-sandbox --remote-debugging-port=9222 --user-data-dir=/tmp/chromium-harvest --disable-gpu "about:blank" &>/dev/null &
-sleep 5
+The script fetches links, deduplicates, fetches each article, downloads images, and **writes to disk itself**:
+- `<英文标题> - <SourceName>原文.md` (full body + images, with a sentinel placeholder for the Chinese title link)
+- `registry.json` (dedup record, written incrementally)
+- `index.md` (daily index, using sentinel titles until finalize)
+- `.manifest.json` (temp — consumed by finalize)
+
+It prints a **compact manifest** to stdout (no full body — only a `bodyPreview` ~4000 chars per article). Read it.
+
+> **Prerequisite** for `browser` sources: Chromium must be running on port 9222:
+> ```bash
+> curl -s http://localhost:9222/json/version 2>/dev/null | python3 -c "import sys,json; d=json.load(sys.stdin); print('OK:', d.get('Browser',''))"
+> ```
+> If not running: `chromium --no-sandbox --remote-debugging-port=9222 --user-data-dir=/tmp/chromium-harvest --disable-gpu "about:blank" &>/dev/null & sleep 5`
+
+#### Step 2. Generate Chinese summaries → write `.summaries.json`
+From the compact manifest, for each article write one entry to `<dayDir>/.summaries.json`:
+
+```json
+[
+  {
+    "url": "https://theconversation.com/...-286430",
+    "titleZh": "特朗普新规或拖慢联邦科研",
+    "category": "科技/AI",
+    "tags": ["科技/AI", "The Conversation"],
+    "summaryZh": "## 事件概述\n...\n\n## 背景\n...\n\n## 关键引语\n> \"...\"\n\n## 影响分析\n..."
+  }
+]
 ```
 
-### 3. Build dedup set
-Scan the last `registryWindowDays` (default 7) days of registry files:
+- `titleZh`: concise Chinese title (≤30 chars)
+- `category` / `tags`: the agent's classification (overrides the script's keyword guess)
+- `summaryZh`: the 4-section markdown body (事件概述 / 背景 / 关键引语 / 影响分析)
+- The `bodyPreview` in the manifest is enough to write the summary. **Do not re-read the full 原文.md** — if you need more, read a targeted slice.
+
+#### Step 3. Run finalize (script stitches summaries in)
+```bash
+node finalize.js "<dayDir>"
 ```
-<vaultPath>/<source.vaultFolder>/YYYYMMDD/registry.json
+This writes each `<中文标题>.md`, patches the sentinel links in the 原文.md files, updates `registry.json` + `index.md` with real titles, and removes `.manifest.json` + `.summaries.json`.
+
+### Flow B — Preview (read-only, no archive)
+
+For "返回最新文章数据" / "看看最新文章" requests where the user does **not** ask to save:
+
+```bash
+node harvest.js <source_id> <count> --preview        # titles + metadata + bodyPreview
+node harvest.js <source_id> <count> --preview --full  # ... + full body
 ```
-Collect all URLs into a Set.
 
-### 4. Fetch homepage links
-- **browser method**: use `scripts/fetcher-browser.js`
-- **direct method**: use `scripts/fetcher-direct.js`
+No files are written. stdout is a compact JSON list (one call, no re-fetching / re-parsing). Use `--full` only when the user explicitly wants the complete article text.
 
-### 5. Filter + slice
-Remove URLs already in dedup set. Take first `count` new ones.
+### Guardrails (token hygiene)
 
-### 6. Process each article
-For each URL (in series, not parallel):
-1. Fetch full article (title, date, authors, body, images)
-2. Download images → `<vaultPath>/<vaultFolder>/YYYYMMDD/assets/<slug>-<n>.jpg`
-3. Generate Chinese summary doc → `<vaultPath>/<vaultFolder>/YYYYMMDD/<中文标题>.md`
-4. Generate original doc → `<vaultPath>/<vaultFolder>/YYYYMMDD/<英文标题> - <SourceName>原文.md`
-5. Update `YYYYMMDD/index.md`
-6. Append to `YYYYMMDD/registry.json`
+- **Don't re-run `harvest.js`** for the same source/day just to "get more" — it deduplicates via the registry, so a second run only fetches genuinely new articles (and merges them into the pending `.manifest.json`, which is safe). If you need more, increase `count` on the first run.
+- **Never `Read` the 原文.md files** to write summaries — the `bodyPreview` in the manifest is sufficient. Targeted slice reads are OK if truly needed.
+- **One `finalize.js` call** after writing `.summaries.json`. Do not hand-write file I/O — the scripts handle all original/registry/index/summary writes.
 
 ---
 
 ## Document Formats
 
-### Chinese Summary Doc
+> These docs are **generated by the scripts**, not hand-written by the agent. Shown here for reference (e.g. when debugging or reading archived files). The agent only writes `.summaries.json`; `finalize.js` produces the Chinese summary docs and patches the originals.
+
+### Original Doc (written by `harvest.js`)
+Filename: `<英文标题> - <SourceName>原文.md`
+
+Between Step 1 (harvest) and Step 3 (finalize), the Chinese-title links in this file are a **sentinel placeholder** `__SUMMARY_<slug>__` — in both the frontmatter `related` field and the footer wikilink. `finalize.js` replaces them with the real `titleZh`. Image captions: only rendered when `alt` text exists (empty alt → bare embed, no `**`).
+
+```markdown
+---
+title: "<英文标题>"
+date: YYYY-MM-DD
+tags: [...]
+source: "<url>"
+publisher: <source>
+authors: "<authors>"
+type: 原文
+related: "[[<中文标题>]]"   # sentinel __SUMMARY_<slug>__ until finalize
+saved: YYYY-MM-DD
+---
+
+# <英文标题>
+
+> [!info] Source
+> **Published:** YYYY-MM-DD | **URL:** <url>
+> **中文摘要：** [[<中文标题>]]
+
+---
+
+## 配图
+![[assets/<slug>-1.jpg]]
+*<alt text — omitted if empty>*
+
+---
+
+## 正文
+<full article body>
+
+---
+[[<中文标题>]]
+```
+
+### Chinese Summary Doc (written by `finalize.js`)
 Filename: `<中文标题>.md` (concise, ≤30 chars)
 
 ```markdown
@@ -125,48 +249,13 @@ saved: YYYY-MM-DD
 [[<英文标题> - <SourceName>原文]]
 ```
 
-### Original Doc
-Filename: `<英文标题> - <SourceName>原文.md`
-
-```markdown
----
-title: "<英文标题>"
-date: YYYY-MM-DD
-tags: [...]
-source: "<url>"
-publisher: <source>
-authors: "<authors>"
-type: 原文
-related: "[[<中文标题>]]"
-saved: YYYY-MM-DD
----
-
-# <英文标题>
-
-> [!info] Source
-> **Published:** YYYY-MM-DD | **URL:** <url>
-> **中文摘要：** [[<中文标题>]]
-
----
-
-## 配图
-![[assets/<slug>-1.jpg]]
-*<alt text>*
-
----
-
-## 正文
-<full article body>
-
----
-[[<中文标题>]]
-```
-
 ---
 
 ## index.md Format
 
-File: `<vaultFolder>/YYYYMMDD/index.md`
+File: `<vaultFolder>/YYYYMMDD/index.md` (written by `harvest.js`, rewritten by `finalize.js`)
+
+Image asset count is a **real file count** of `assets/` (not "articles with images"), so it reflects downloaded files even before finalize.
 
 ```markdown
 ---
@@ -201,7 +290,9 @@ last_updated: "<ISO timestamp>"
 
 ## registry.json Format
 
-File: `<vaultFolder>/YYYYMMDD/registry.json`
+File: `<vaultFolder>/YYYYMMDD/registry.json` (written incrementally by `harvest.js`, patched by `finalize.js`)
+
+`titleZh` is the sentinel `__SUMMARY_<slug>__` until finalize supplies the real title.
 
 ```json
 {
@@ -212,9 +303,11 @@ File: `<vaultFolder>/YYYYMMDD/registry.json`
       "url": "https://...",
       "titleZh": "中文标题",
       "titleEn": "English Title",
-      "processedAt": "2026-02-28T14:30:00+08:00",
+      "processedAt": "2026-07-24T14:30:00.000Z",
       "category": "地缘政治",
-      "tags": ["tag1", "tag2"]
+      "tags": ["tag1", "tag2"],
+      "slug": "theconversation-com-some-article-286430",
+      "originalFilename": "Some Article - The Conversation原文.md"
     }
   ]
 }

scripts/fetcher-browser.js

@@ -7,7 +7,9 @@
 const http = require('http');
 const WebSocket = require('ws');
 
-const CDP_PORT = parseInt(process.env.CDP_PORT_OVERRIDE || '9223');
+// Chromium CDP debug port — must match the --remote-debugging-port used to start
+// Chromium (9222 per SKILL.md / README). Override via env if you run elsewhere.
+const CDP_PORT = parseInt(process.env.CDP_PORT_OVERRIDE || '9222');
 
 function getWsUrl() {
   return new Promise((resolve, reject) => {

scripts/fetcher-direct.js

@@ -30,6 +30,45 @@ function fetchHtml(url) {
   });
 }
 
+// Extract author names from JSON-LD, then <meta name="author"> (one tag per
+// author, as on The Conversation), then rel="author" links. Returns '' if none.
+function extractAuthors(html) {
+  // 1. JSON-LD author (Reuters/Guardian embed ld+json blocks)
+  const ldBlocks = [...html.matchAll(/<script[^>]*type=["']application\/ld\+json["'][^>]*>([\s\S]*?)<\/script>/gi)];
+  for (const b of ldBlocks) {
+    try {
+      const data = JSON.parse(b[1]);
+      const candidates = [];
+      const collect = (obj) => {
+        if (!obj) return;
+        if (Array.isArray(obj)) { obj.forEach(collect); return; }
+        candidates.push(obj);
+        if (obj['@graph']) collect(obj['@graph']);
+      };
+      collect(data);
+      for (const obj of candidates) {
+        let author = obj.author || (obj.mainEntity && obj.mainEntity.author);
+        if (author) {
+          const names = (Array.isArray(author) ? author : [author])
+            .map(a => typeof a === 'string' ? a : (a && a.name) || '')
+            .filter(Boolean);
+          if (names.length) return [...new Set(names)].slice(0, 3).join(', ');
+        }
+      }
+    } catch (e) {}
+  }
+  // 2. <meta name="author" content="..."> — multiple tags for co-authors
+  const metaMatches = [...html.matchAll(/<meta[^>]+name=["']author["'][^>]+content=["']([^"']+)["']/gi)];
+  if (metaMatches.length) {
+    const names = [...new Set(metaMatches.map(m => m[1].trim()).filter(Boolean))];
+    if (names.length) return names.slice(0, 3).join(', ');
+  }
+  // 3. rel="author" link text
+  const relMatch = html.match(/<a[^>]+rel=["']author["'][^>]*>([^<]+)<\/a>/i);
+  if (relMatch && relMatch[1].trim()) return relMatch[1].trim();
+  return '';
+}
+
 function extractArticle(html, url) {
   // Title
   const titleMatch = html.match(/<title[^>]*>(.*?)<\/title>/is);
@@ -42,16 +81,17 @@ function extractArticle(html, url) {
   const date = dateMatch ? dateMatch[1] : '';
 
   // Authors
-  const authorMatches = html.match(/["']name["']\s*:\s*["']([^"']{3,60})["']/g) || [];
-  const authors = [...new Set(authorMatches.map(m => m.match(/["']name["']\s*:\s*["']([^"']+)["']/)[1])
-    .filter(a => a.length < 60 && !a.includes('{')))].slice(0, 3).join(', ');
+  const authors = extractAuthors(html);
 
-  // Images
+  // Images — skip logos/icons and tracking pixels (e.g. facebook.com/tr), dedupe by src
   const imgMatches = [...html.matchAll(/<img[^>]+src=["'](https?:\/\/[^"']+)["'][^>]*>/gi)];
+  const seenSrc = new Set();
   const imgs = imgMatches.map(m => {
     const altMatch = m[0].match(/alt=["']([^"']*)["']/i);
     return { src: m[1], alt: altMatch ? altMatch[1] : '', width: 500 };
-  }).filter(i => !i.src.includes('logo') && !i.src.includes('icon') && i.src.length > 50).slice(0, 6);
+  }).filter(i => !i.src.includes('logo') && !i.src.includes('icon')
+    && !i.src.includes('facebook.com/tr') && i.src.length > 50
+    && !seenSrc.has(i.src) && seenSrc.add(i.src)).slice(0, 6);
 
   // Article body - try structured extraction
   let body = '';
@@ -66,12 +106,17 @@ function extractArticle(html, url) {
     // Extract all paragraphs, clean HTML
     const paraMatches = [...html.matchAll(/<p[^>]*>(.*?)<\/p>/gis)];
     const noise = ['Sign up', 'Newsletter', 'Subscribe', 'Follow us', 'Read more', 'Click here',
-      'opens new tab', 'Thomson Reuters', 'All quotes delayed', 'Access unmatched'];
+      'opens new tab', 'Thomson Reuters', 'All quotes delayed', 'Access unmatched',
+      'Copy link', 'Bluesky', 'Write an article and join'];
     body = paraMatches.map(m => {
       let text = m[1].replace(/<[^>]+>/g, '').replace(/&nbsp;/g, ' ')
         .replace(/&amp;/g, '&').replace(/&lt;/g, '<').replace(/&gt;/g, '>').trim();
       return text;
-    }).filter(t => t.length > 60 && !noise.some(n => t.includes(n))).join('\n\n');
+    }).filter(t => {
+      const words = t.replace(/\s+/g, ' ').trim().split(/\s+/);
+      // drop share-button blobs / fragments (much whitespace, few real words)
+      return words.length >= 8 && t.length > 60 && !noise.some(n => t.includes(n));
+    }).join('\n\n');
   }
 
   return { title, date, authors, imgs, body: body.substring(0, 15000) };

scripts/finalize.js

+#!/usr/bin/env node
+/**
+ * finalize.js - Patch Chinese summaries into an already-archived day directory.
+ * Usage: node finalize.js <dayDir>
+ *
+ * Reads:
+ *   <dayDir>/.manifest.json   (written by harvest.js — maps url → slug → original filename)
+ *   <dayDir>/.summaries.json  (written by the agent — Chinese title + 4-section summary)
+ *
+ * For each article with a summary:
+ *   1. Write <dayDir>/<中文标题>.md           (the Chinese summary doc)
+ *   2. Patch the sentinel __SUMMARY_<slug>__ in the original 原文.md     (related + footer wikilink)
+ *   3. Update the registry entry (titleZh / category / tags)
+ * Finally rewrites index.md with real titles and removes the two temp files.
+ *
+ * Missing summary for an article → warn and leave its sentinel in place (non-fatal).
+ * Missing .summaries.json → error and exit.
+ */
+
+const fs = require('fs');
+const path = require('path');
+
+const {
+  generateSummaryDoc,
+  updateIndex,
+  sentinelFor,
+  originalFilename
+} = require('./harvest');
+
+function loadJson(p, label) {
+  if (!fs.existsSync(p)) {
+    console.error(`✗ ${label} not found: ${p}`);
+    process.exit(1);
+  }
+  try {
+    return JSON.parse(fs.readFileSync(p, 'utf8'));
+  } catch (e) {
+    console.error(`✗ Failed to parse ${label}: ${e.message}`);
+    process.exit(1);
+  }
+}
+
+function printHelp() {
+  console.log(`finalize.js — Patch Chinese summaries into an archived day directory.
+
+Usage:
+  node finalize.js <dayDir>
+
+Arguments:
+  dayDir   Path to the YYYYMMDD directory produced by harvest.js
+           (e.g. /path/to/vault/The Conversation/20260724)
+
+Prerequisites:
+  - <dayDir>/.manifest.json   (written by harvest.js)
+  - <dayDir>/.summaries.json  (written by the agent — Chinese title + 4-section summary)
+
+What it does:
+  1. Writes <dayDir>/<中文标题>.md for each article with a summary
+  2. Replaces sentinel placeholders in the 原文.md files with real titles
+  3. Updates registry.json and index.md with real titles/categories/tags
+  4. Removes .manifest.json and .summaries.json
+
+Options:
+  -h, --help   Show this help message
+
+Example:
+  node finalize.js "/Users/me/obsidian/Harvester/The Conversation/20260724"`);
+}
+
+function main() {
+  const args = process.argv.slice(2);
+  if (args.includes('-h') || args.includes('--help')) {
+    printHelp();
+    process.exit(0);
+  }
+
+  const dayDir = args.find(a => !a.startsWith('-'));
+  if (!dayDir) {
+    console.error('Usage: node finalize.js <dayDir>');
+    console.error('Example: node finalize.js /path/to/vault/The Conversation/20260724');
+    console.error('Run "node finalize.js --help" for details.');
+    process.exit(1);
+  }
+  if (!fs.existsSync(dayDir) || !fs.statSync(dayDir).isDirectory()) {
+    console.error(`✗ Day directory does not exist: ${dayDir}`);
+    process.exit(1);
+  }
+
+  const manifestPath = path.join(dayDir, '.manifest.json');
+  const summariesPath = path.join(dayDir, '.summaries.json');
+
+  const manifest = loadJson(manifestPath, '.manifest.json');
+  const summaries = loadJson(summariesPath, '.summaries.json');
+
+  const sourceName = manifest.sourceName || 'Unknown';
+  // Index summaries by url for O(1) lookup
+  const summaryByUrl = new Map();
+  for (const s of (Array.isArray(summaries) ? summaries : [])) {
+    if (s && s.url) summaryByUrl.set(s.url, s);
+  }
+
+  console.error(`\n🖌️  Finalizing ${dayDir}`);
+  console.error(`   Articles in manifest: ${manifest.articles.length}`);
+  console.error(`   Summaries provided:   ${summaryByUrl.size}`);
+
+  // Build the registry path (same convention as harvest.js)
+  const regPath = path.join(dayDir, 'registry.json');
+  let registry = { date: path.basename(dayDir), source: manifest.sourceId, articles: [] };
+  if (fs.existsSync(regPath)) {
+    try { registry = JSON.parse(fs.readFileSync(regPath, 'utf8')); } catch (e) {}
+  }
+
+  let patched = 0, skipped = 0;
+
+  for (const article of manifest.articles) {
+    const { url, slug, titleEn, originalFilename: origName } = article;
+    const summary = summaryByUrl.get(url);
+
+    if (!summary) {
+      console.error(`  ⚠️  No summary for: ${titleEn} — leaving sentinel in place`);
+      skipped++;
+      continue;
+    }
+
+    const titleZh = summary.titleZh;
+    if (!titleZh || typeof titleZh !== 'string' || !titleZh.trim()) {
+      console.error(`  ⚠️  Empty titleZh for: ${titleEn} — skipping`);
+      skipped++;
+      continue;
+    }
+
+    const sentinel = sentinelFor(slug);
+    const summaryFilePath = path.join(dayDir, `${titleZh}.md`);
+    const origFilePath = path.join(dayDir, origName || originalFilename(titleEn, sourceName));
+
+    // 1. Write the Chinese summary doc
+    const summaryDoc = generateSummaryDoc({
+      titleZh,
+      titleEn,
+      date: article.date,
+      authors: article.authors,
+      source: url,
+      category: summary.category || article.category,
+      tags: summary.tags || article.tags,
+      summaryZh: summary.summaryZh
+    }, sourceName);
+    fs.writeFileSync(summaryFilePath, summaryDoc);
+
+    // 2. Patch the sentinel in the original doc (frontmatter related + footer wikilink)
+    if (fs.existsSync(origFilePath)) {
+      const orig = fs.readFileSync(origFilePath, 'utf8');
+      // Replace all occurrences of the sentinel with the real title.
+      const patchedOrig = orig.split(sentinel).join(titleZh);
+      fs.writeFileSync(origFilePath, patchedOrig);
+    } else {
+      console.error(`  ⚠️  Original doc missing: ${origName} — summary written but link unpatched`);
+    }
+
+    // 3. Update the registry entry for this url
+    const regEntry = registry.articles.find(a => a.url === url);
+    if (regEntry) {
+      regEntry.titleZh = titleZh;
+      if (summary.category) regEntry.category = summary.category;
+      if (summary.tags) regEntry.tags = summary.tags;
+    }
+
+    patched++;
+    console.error(`  ✓ ${titleEn} → ${titleZh}`);
+  }
+
+  // Persist the patched registry
+  fs.writeFileSync(regPath, JSON.stringify(registry, null, 2));
+
+  // 4. Rewrite index.md using real titles (from the now-updated registry).
+  //    Index links resolve to summary docs where a titleZh exists, else sentinel.
+  const indexArticles = registry.articles.map(a => ({
+    titleZh: a.titleZh,
+    date: a.date || '',
+    category: a.category || '国际',
+    tags: a.tags || []
+  }));
+  updateIndex(dayDir, indexArticles, sourceName);
+
+  // 5. Cleanup temp files
+  try { fs.unlinkSync(manifestPath); } catch (e) {}
+  try { fs.unlinkSync(summariesPath); } catch (e) {}
+
+  console.error(`\n✅ Done: ${patched} finalized, ${skipped} skipped`);
+  console.error(`   Registry + index updated, temp files removed`);
+}
+
+if (require.main === module) {
+  main();
+}

scripts/harvest.js

 #!/usr/bin/env node
 /**
  * harvest.js - Main entry point for news harvesting
- * Usage: node harvest.js <source_id> <count>
- * Example: node harvest.js reuters 5
+ * Usage:
+ *   node harvest.js <source_id> <count>              # archive: fetch + write originals/registry/index
+ *   node harvest.js <source_id> <count> --preview    # read-only: print article list (no writes)
+ *   node harvest.js <source_id> <count> --preview --full  # read-only + full body
+ *
+ * Archive flow (writes everything deterministic to disk itself):
+ *   fetch links → dedup → fetch each article → download images → write 原文.md +
+ *   registry.json + index.md (using sentinel placeholder for the Chinese title) →
+ *   print a COMPACT manifest (no full body) + write .manifest.json.
+ * The agent then reads the manifest, writes .summaries.json, and runs finalize.js
+ * to generate the Chinese summary docs and patch the sentinel links.
  */
 
 const fs = require('fs');
@@ -14,6 +23,9 @@ const SKILL_DIR = path.dirname(__dirname);
 const CONFIG_PATH = path.join(SKILL_DIR, 'config.json');
 const TEMPLATE_PATH = path.join(SKILL_DIR, 'references', 'config-template.json');
 
+// Body preview length sent to the agent (full body stays on disk, never in context)
+const PREVIEW_LEN = 4000;
+
 // ── Helpers ─────────────────────────────────────────────────────────────────
 
 function loadConfig() {
@@ -36,6 +48,10 @@ function todayStr() {
   return `${d.getFullYear()}${String(d.getMonth()+1).padStart(2,'0')}${String(d.getDate()).padStart(2,'0')}`;
 }
 
+function dateShort(dateStr) {
+  return (dateStr || '').substring(0, 10) || todayStr().replace(/(\d{4})(\d{2})(\d{2})/, '$1-$2-$3');
+}
+
 function getCategory(url, title) {
   const u = url.toLowerCase(), t = title.toLowerCase();
   if (/asia-pacific|pakistan|afghanistan/.test(u) || /巴基斯坦|阿富汗/.test(t)) return '地缘政治';
@@ -57,6 +73,16 @@ function slugFromUrl(url) {
     .replace(/[^a-z0-9-]/gi, '-').replace(/-+/g, '-').substring(0, 60);
 }
 
+// Placeholder Chinese title used in 原文.md links, registry and index until the
+// agent supplies a real translation. finalize.js replaces it with the real title.
+function sentinelFor(slug) {
+  return `__SUMMARY_${slug}__`;
+}
+
+function originalFilename(titleEn, sourceName) {
+  return `${titleEn} - ${sourceName}原文.md`;
+}
+
 // ── Registry ─────────────────────────────────────────────────────────────────
 
 function buildDedupSet(config, sourceId) {
@@ -117,15 +143,14 @@ function downloadImage(url, dest) {
 
 // ── Document Generation ───────────────────────────────────────────────────────
 
-function generateSummaryDoc(data, originPath, sourceName) {
+function generateSummaryDoc(data, sourceName) {
   const { titleZh, titleEn, date, authors, source: url, category, tags } = data;
-  const dateShort = (date || '').substring(0, 10) || todayStr().replace(/(\d{4})(\d{2})(\d{2})/, '$1-$2-$3');
   const tagsYaml = (tags || []).map(t => `  - ${t}`).join('\n');
-  const origName = `${titleEn} - ${sourceName}原文`;
+  const origName = originalFilename(titleEn, sourceName);
 
   return `---
 title: "${titleZh}"
-date: ${dateShort}
+date: ${dateShort(date)}
 tags:
 ${tagsYaml}
 source: "${url}"
@@ -139,7 +164,7 @@ saved: ${todayStr().replace(/(\d{4})(\d{2})(\d{2})/, '$1-$2-$3')}
 # ${titleZh}
 
 > [!info] 文章信息
-> **发布：** ${dateShort} | **来源：** ${sourceName}
+> **发布：** ${dateShort(date)} | **来源：** ${sourceName}
 > **作者：** ${authors || '—'}
 > **原文：** [[${origName}]]
 
@@ -155,15 +180,19 @@ ${data.summaryZh || '（摘要生成中）'}
 
 function generateOriginalDoc(data, sourceName, imgFiles) {
   const { titleEn, titleZh, date, authors, source: url, body } = data;
-  const dateShort = (date || '').substring(0, 10) || todayStr().replace(/(\d{4})(\d{2})(\d{2})/, '$1-$2-$3');
   const tags = (data.tags || []).map(t => `  - ${t}`).join('\n');
+  // Empty-alt guard: only render the caption line when alt text exists, so we
+  // never emit a bare `**` (empty bold) for avatar/tracking images.
   const imgSection = imgFiles.length > 0
-    ? imgFiles.map((f, i) => `![[assets/${f.filename}]]\n*${f.alt || ''}*`).join('\n\n')
+    ? imgFiles.map(f => {
+        const embed = `![[assets/${f.filename}]]`;
+        return f.alt ? `${embed}\n*${f.alt}*` : embed;
+      }).join('\n\n')
     : '（无图片）';
 
   return `---
 title: "${titleEn}"
-date: ${dateShort}
+date: ${dateShort(date)}
 tags:
 ${tags}
 source: "${url}"
@@ -177,7 +206,7 @@ saved: ${todayStr().replace(/(\d{4})(\d{2})(\d{2})/, '$1-$2-$3')}
 # ${titleEn}
 
 > [!info] Source
-> **Published:** ${dateShort} | **URL:** ${url}
+> **Published:** ${dateShort(date)} | **URL:** ${url}
 > **中文摘要：** [[${titleZh}]]
 
 ---
@@ -200,17 +229,27 @@ ${body || '（正文获取失败）'}
 
 // ── index.md Update ───────────────────────────────────────────────────────────
 
-function updateIndex(config, source, today, articles) {
-  const indexPath = path.join(config.vaultPath, source.vaultFolder, today, 'index.md');
+function countAssets(dayDir) {
+  const assetsDir = path.join(dayDir, 'assets');
+  if (!fs.existsSync(assetsDir)) return 0;
+  try {
+    return fs.readdirSync(assetsDir).filter(n => fs.statSync(path.join(assetsDir, n)).isFile()).length;
+  } catch (e) { return 0; }
+}
+
+function updateIndex(dayDir, articles, sourceName) {
+  const indexPath = path.join(dayDir, 'index.md');
+  const today = path.basename(dayDir);
   const dateFormatted = today.replace(/(\d{4})(\d{2})(\d{2})/, '$1-$2-$3');
   const now = new Date().toISOString().replace('T', ' ').substring(0, 16) + ' CST';
+  const assetCount = countAssets(dayDir);
 
   // Category stats
   const catCount = {};
   articles.forEach(a => { catCount[a.category] = (catCount[a.category] || 0) + 1; });
   const catRows = Object.entries(catCount).map(([c, n]) => `| ${CATEGORY_EMOJI[c] || '📰'} ${c} | ${n} |`).join('\n');
   const tableRows = articles.map((a, i) =>
-    `| ${i+1} | [[${a.titleZh}]] | ${a.category} | ${(a.tags||[]).slice(0,3).join(', ')} | ${(a.date||'').substring(0,10)} | ✅ |`
+    `| ${i+1} | [[${a.titleZh}]] | ${a.category} | ${(a.tags||[]).slice(0,3).join(', ')} | ${dateShort(a.date)} | ✅ |`
   ).join('\n');
 
   const catNav = Object.entries(catCount).map(([cat, n]) => {
@@ -221,15 +260,15 @@ function updateIndex(config, source, today, articles) {
   const content = `---
 date: ${dateFormatted}
 type: daily-index
-source: ${source.name}
+source: ${sourceName}
 total_articles: ${articles.length}
 last_updated: "${new Date().toISOString()}"
 ---
 
-# ${source.name} Daily Index · ${dateFormatted}
+# ${sourceName} Daily Index · ${dateFormatted}
 
 > **最近更新：** ${now}
-> **当日归档：** ${articles.length} 篇 | **图片资产：** ${articles.filter(a => a.hasImg).length} 张
+> **当日归档：** ${articles.length} 篇 | **图片资产：** ${assetCount} 张
 > **涵盖分类：** ${Object.keys(catCount).join(' · ')}
 
 ---
@@ -261,12 +300,46 @@ ${catNav}
 // ── Main ──────────────────────────────────────────────────────────────────────
 
 async function main() {
-  const sourceId = process.argv[2];
-  const count = parseInt(process.argv[3] || '5');
+  // Parse args: <source_id> [count] [--preview] [--full] [-h|--help]
+  const argv = process.argv.slice(2);
+  if (argv.includes('-h') || argv.includes('--help')) {
+    console.log(`harvest.js — Fetch and archive news articles from a configured source.
+
+Usage:
+  node harvest.js <source_id> [count] [--preview] [--full]
+
+Arguments:
+  source_id   A source id from config.json (e.g. the-conversation, reuters)
+  count       Max number of new articles to fetch (default 5)
+
+Modes:
+  (default)     Archive: fetch + write 原文.md / registry / index to vault,
+                print compact manifest (bodyPreview only, no full body)
+  --preview     Read-only: print article list (titles + metadata + bodyPreview),
+                no files written
+  --preview --full   Read-only + include full article body in output
+
+Options:
+  -h, --help   Show this help message
+
+Examples:
+  node harvest.js the-conversation 3
+  node harvest.js the-conversation 3 --preview
+  node harvest.js the-conversation 3 --preview --full
+  node harvest.js reuters 5`);
+    process.exit(0);
+  }
+  const preview = argv.includes('--preview');
+  const full = argv.includes('--full');
+  const positional = argv.filter(a => !a.startsWith('-'));
+  const sourceId = positional[0];
+  const count = parseInt(positional[1] || '5');
 
   if (!sourceId) {
-    console.error('Usage: node harvest.js <source_id> <count>');
+    console.error('Usage: node harvest.js <source_id> [count] [--preview] [--full]');
     console.error('Example: node harvest.js reuters 5');
+    console.error('         node harvest.js the-conversation 3 --preview --full');
+    console.error('Run "node harvest.js --help" for details.');
     process.exit(1);
   }
 
@@ -283,17 +356,17 @@ async function main() {
     process.exit(1);
   }
 
-  console.error(`\n🗞️  Harvesting ${source.name} (max ${count} articles)`);
+  const mode = preview ? (full ? 'preview (full)' : 'preview') : 'archive';
+  console.error(`\n🗞️  Harvesting ${source.name} (max ${count} articles) — ${mode}`);
   console.error(`   Method: ${source.method}`);
 
-  // Load fetcher
   const fetcher = source.method === 'browser'
     ? require('./fetcher-browser')
     : require('./fetcher-direct');
 
-  // Build dedup set
-  const seen = buildDedupSet(config, sourceId);
-  console.error(`   Registry: ${seen.size} URLs in last ${config.registryWindowDays} days`);
+  // Build dedup set (skip in preview — preview shows latest regardless)
+  const seen = preview ? new Set() : buildDedupSet(config, sourceId);
+  console.error(`   Registry: ${seen.size} URLs in last ${config.registryWindowDays || 7} days`);
 
   // Fetch homepage links
   console.error(`\n📡 Fetching homepage: ${source.homepageUrl}`);
@@ -306,27 +379,68 @@ async function main() {
   }
   console.error(`   Found ${links.length} article links`);
 
-  // Normalize URLs (strip query params for dedup, e.g. Bloomberg ?srnd=)
+  // Normalize URLs (strip query params for dedup)
   const normalizeUrl = (u) => { try { const p = new URL(u); p.search = ''; return p.toString(); } catch(e) { return u; } };
   const normalizedLinks = links.map(u => normalizeUrl(u));
   const newLinks = normalizedLinks.filter(u => !seen.has(u)).slice(0, count);
   const skipped = normalizedLinks.filter(u => seen.has(u)).length;
   console.error(`   New: ${newLinks.length} | Already seen: ${skipped}`);
 
+  // ── Preview mode: fetch only, print compact list, write nothing ───────────
+  if (preview) {
+    const out = [];
+    for (let i = 0; i < newLinks.length; i++) {
+      const url = newLinks[i];
+      console.error(`\n[${i+1}/${newLinks.length}] ${url}`);
+      let articleData;
+      try {
+        articleData = await fetcher.fetchPage(url);
+      } catch (e) {
+        console.error(`  ✗ Fetch failed: ${e.message}`);
+        continue;
+      }
+      if (!articleData || !articleData.title) { console.error('  ✗ No title, skip'); continue; }
+      const category = getCategory(url, articleData.title);
+      const entry = {
+        title: articleData.title,
+        date: articleData.date,
+        authors: articleData.authors,
+        url,
+        category,
+        charCount: (articleData.body || '').length,
+        imgCount: (articleData.imgs || []).length,
+        imgs: (articleData.imgs || []).map(im => ({ alt: im.alt || '', src: im.src }))
+      };
+      if (full) {
+        entry.body = articleData.body || '';
+      } else {
+        entry.bodyPreview = (articleData.body || '').substring(0, PREVIEW_LEN);
+      }
+      out.push(entry);
+      console.error(`  ✓ ${articleData.title} (${entry.charCount} chars, ${entry.imgCount} imgs)`);
+    }
+    process.stdout.write(JSON.stringify({
+      sourceId, sourceName: source.name, preview: true, full, count: out.length, articles: out
+    }, null, 2));
+    console.error(`\n✅ Preview: ${out.length} articles (no files written)`);
+    return;
+  }
+
+  // ── Archive mode: fetch + write originals/registry/index ─────────────────
   if (newLinks.length === 0) {
     console.error('No new articles to process.');
-    const manifest = { sourceId, sourceName: source.name, today: todayStr(), processed: [], skipped, newCount: 0 };
-    process.stdout.write(JSON.stringify(manifest, null, 2));
+    process.stdout.write(JSON.stringify({
+      sourceId, sourceName: source.name, today: todayStr(), dayDir: null,
+      summariesPath: null, articles: [], skipped, newCount: 0
+    }, null, 2));
     return;
   }
 
-  // Setup dirs
   const today = todayStr();
   const dayDir = path.join(config.vaultPath, source.vaultFolder, today);
   const assetsDir = path.join(dayDir, 'assets');
   fs.mkdirSync(assetsDir, { recursive: true });
 
-  // Process articles
   const processed = [];
 
   for (let i = 0; i < newLinks.length; i++) {
@@ -341,7 +455,7 @@ async function main() {
       continue;
     }
 
-    if (!articleData || !articleData.title || articleData.body.length < 100) {
+    if (!articleData || !articleData.title || (articleData.body || '').length < 100) {
       console.error(`  ✗ Insufficient content (body: ${(articleData?.body||'').length} chars)`);
       continue;
     }
@@ -349,7 +463,9 @@ async function main() {
     const { title: titleEn, date, authors, imgs = [], body } = articleData;
     const slug = slugFromUrl(url);
     const category = getCategory(url, titleEn);
-    const tags = [category, source.name, ...(date || '').substring(0,4) ? [] : []].filter(Boolean);
+    const tags = [category, source.name];
+    const titleZh = sentinelFor(slug);          // patched by finalize.js
+    const origName = originalFilename(titleEn, source.name);
 
     // Download images
     const imgFiles = [];
@@ -362,45 +478,102 @@ async function main() {
       if (ok) imgFiles.push({ filename, alt: imgs[j].alt, src: imgSrc });
     }
 
-    // Placeholder for Chinese title (agent will generate proper translation)
-    // For now use a truncated cleaned version that the agent will improve
-    const titleZh = `[需翻译] ${titleEn.substring(0, 40)}`;
-
     const articleInfo = {
-      url,
-      titleEn,
-      titleZh,
-      date,
-      authors,
-      body,
-      imgs: imgFiles,
-      category,
-      tags,
-      hasImg: imgFiles.length > 0,
-      slug,
-      dayDir,
-      sourceName: source.name
+      url, titleEn, titleZh, date, authors, body,
+      imgs: imgFiles, category, tags, slug,
+      originalFilename: origName, sourceName: source.name
     };
 
+    // Write original doc (full body on disk, never sent to the agent)
+    fs.writeFileSync(path.join(dayDir, origName), generateOriginalDoc(articleInfo, source.name, imgFiles));
+
+    // Append to registry (dedup record, written immediately so a crash keeps it)
+    appendRegistry(config, source, today, {
+      url, titleZh, titleEn, processedAt: new Date().toISOString(),
+      category, tags, slug, originalFilename: origName
+    });
+
+    // Refresh daily index from the running set (sentinel titles until finalize)
+    updateIndex(dayDir, [...processed, articleInfo], source.name);
+
     processed.push(articleInfo);
-    console.error(`  ✓ ${titleEn} (${imgFiles.length} imgs, ${body.length} chars)`);
+    console.error(`  ✓ ${titleEn} (${imgFiles.length} imgs, ${body.length} chars) → ${origName}`);
     await sleep(1000);
   }
 
-  // Output manifest for agent to use
+  // Compact manifest for the agent: NO full body, only a preview slice.
+  const newArticles = processed.map(a => ({
+    url: a.url,
+    titleEn: a.titleEn,
+    date: a.date,
+    authors: a.authors,
+    category: a.category,
+    tags: a.tags,
+    slug: a.slug,
+    originalFilename: a.originalFilename,
+    charCount: (a.body || '').length,
+    imgs: a.imgs.map(im => ({ filename: im.filename, alt: im.alt || '', src: im.src })),
+    bodyPreview: (a.body || '').substring(0, PREVIEW_LEN)
+  }));
+
+  // Merge into any existing .manifest.json rather than overwriting it. A second
+  // harvest run in the same day dir (or recovery after a crash) must not lose
+  // the slug→filename mapping for articles already on disk but not yet finalized.
+  // finalize.js consumes the merged set so no sentinel is left dangling.
+  const manifestPath = path.join(dayDir, '.manifest.json');
+  let mergedArticles = newArticles;
+  if (fs.existsSync(manifestPath)) {
+    try {
+      const prev = JSON.parse(fs.readFileSync(manifestPath, 'utf8'));
+      const prevArticles = Array.isArray(prev.articles) ? prev.articles : [];
+      const seenUrl = new Set(newArticles.map(a => a.url));
+      mergedArticles = [...newArticles, ...prevArticles.filter(a => !seenUrl.has(a.url))];
+    } catch (e) {
+      console.error(`  ⚠️  Could not merge previous .manifest.json (${e.message}); overwriting`);
+    }
+  }
+
   const manifest = {
     sourceId,
     sourceName: source.name,
     today,
     dayDir,
-    processed,
+    summariesPath: path.join(dayDir, '.summaries.json'),
     skipped,
-    newCount: processed.length
+    newCount: processed.length,
+    nextStep: `Read this manifest, write Chinese summaries to ${path.join(dayDir, '.summaries.json')}, then run: node scripts/finalize.js "${dayDir}"`,
+    summariesContract: {
+      path: path.join(dayDir, '.summaries.json'),
+      format: 'array of { url, titleZh, category, tags[], summaryZh } — one per article url; summaryZh = the 4 sections markdown (事件概述/背景/关键引语/影响分析)'
+    },
+    articles: mergedArticles
   };
 
-  process.stdout.write(JSON.stringify(manifest, null, 2));
-  console.error(`\n✅ Done: ${processed.length} processed, ${skipped} skipped`);
-  console.error('   Agent should now: generate Chinese summaries, create docs, update registry');
+  // Persist the merged manifest so finalize.js can map every url → slug → original filename
+  fs.writeFileSync(manifestPath, JSON.stringify(manifest, null, 2));
+
+  // stdout shows only THIS run's new articles (the agent summarizes what it just fetched)
+  process.stdout.write(JSON.stringify({ ...manifest, articles: newArticles }, null, 2));
+  console.error(`\n✅ Done: ${processed.length} processed this run, ${skipped} skipped`);
+  console.error(`   Manifest (${mergedArticles.length} total pending): ${manifestPath}`);
+  console.error(`   Next: generate .summaries.json, then run finalize.js`);
 }
 
-main().catch(e => { console.error('Fatal:', e); process.exit(1); });
+// Run only when invoked directly (not when required by finalize.js)
+if (require.main === module) {
+  main().catch(e => { console.error('Fatal:', e); process.exit(1); });
+}
+
+// Exported for finalize.js
+module.exports = {
+  generateSummaryDoc,
+  generateOriginalDoc,
+  updateIndex,
+  appendRegistry,
+  getCategory,
+  CATEGORY_EMOJI,
+  slugFromUrl,
+  sentinelFor,
+  originalFilename,
+  dateShort
+};