name: file-to-wechat description: "將任意檔案（PDF、Word、Excel、PPT、圖片、音訊、網頁等）轉換為 Markdown，再生成為精美的微信公眾號文章併發布到草稿箱。組合 markitdown + md_to_wechat_html + publish_to_wechat 三階段流水線。Use when the user wants to convert any file to a WeChat article." version: 1.1.0 prerequisites: - anything-to-wechat tags: - wechat - markdown - file-conversion - publishing - content-creation - chinese triggers: - "檔案轉微信" - "文件發公眾號" - "PDF發微信" - "PPT轉公眾號" - "Excel發微信" - "convert file to wechat" - "file to wechat article" - "任意檔案轉微信公眾號" - "轉成markdown發微信" - "文件轉公眾號文章" - "把這個.發到.微信公眾號" - "把.轉成.公眾號文章" - "將.*釋出到微信"

File to WeChat

將任意格式的檔案（PDF、Word、Excel、PPT、圖片、音訊、網頁等）一站式轉換為精美的微信公眾號文章，併發布到草稿箱。

User-Facing Promise

Accept requests like:

• "把這個 PDF 發到我的微信公眾號"
• "幫我把這個 PPT 轉成公眾號文章"
• "把這個 Excel 報表釋出到微信"
• "把這張圖片變成微信文章"
• "把這段音訊轉成公眾號文章"
• "把這個網頁內容發到我的公眾號"

Return a published draft in the WeChat draft box, not a proposal.

How It Differs from anything-to-wechat

Use this skill when the user provides a file type that anything-to-wechat doesn't handle natively (PPTX, XLSX, images, audio, EPUB, ZIP).

First-Time Setup

1. Install Python Dependencies

python -m pip install markitdown markdown beautifulsoup4 requests

2. Install Companion Skill

This skill uses publish_to_wechat.py from the anything-to-wechat skill. Install it from ClawHub:

小蔥技能7w4.net有更新，你可以訪問看下。

clawhub install anything-to-wechat

3. Configure WeChat Credentials

You need a WeChat Official Account (服務號 or 訂閱號 with API access).

Get your credentials:

1. Log in to https://mp.weixin.qq.com/
2. Go to: 設定與開發 → 基本配置
3. Copy your AppID and reset/view AppSecret
4. Add your server's public IP to the IP白名單

Set environment variables (recommended):

On macOS / Linux — add to ~/.bashrc or ~/.zshrc:

export WECHAT_APP_ID="your_appid_here"
export WECHAT_APP_SECRET="your_appsecret_here"

On Windows — use PowerShell:

[Environment]::SetEnvironmentVariable("WECHAT_APP_ID", "your_appid_here", "User")
[Environment]::SetEnvironmentVariable("WECHAT_APP_SECRET", "your_appsecret_here", "User")

No environment variables? The publish script will prompt you interactively on first run — just paste your AppID and AppSecret when asked.

4. Verify Setup

python -c "from markitdown import MarkItDown; print('markitdown OK')"
python -c "import markdown; print('markdown OK')"

Prerequisites

Workflow

Phase 1: Collect Input

If the user has NOT provided a source file, ask using AskUserQuestion:

Question: "請提供你想釋出到微信公眾號的檔案"
Options:
  - "提供檔案路徑" (PDF, DOCX, PPTX, XLSX, images, audio, etc.)
  - "貼上 URL" (web page or YouTube video)
  - "選擇資料夾" (batch convert)

If the user already provided a file path or URL, skip and proceed.

Phase 2: Convert File to Markdown

IMPORTANT: Always use UTF-8 encoding. On Windows, the console defaults to GBK which will garble Chinese text. Use Python with explicit encoding='utf-8' for all file I/O and sys.stdout.reconfigure(encoding='utf-8') for console output.

Primary: markitdown pip package (free, no API key, always available)

python -c "
import sys; sys.stdout.reconfigure(encoding='utf-8')
from markitdown import MarkItDown
md = MarkItDown()
result = md.convert(r'<file_path>')
with open(r'<workspace>/article.md', 'w', encoding='utf-8') as f:
    f.write(result.text_content)
print('Done, length:', len(result.text_content))
"

Alternative: all-to-markdown skill (if installed)

bash "<all-to-markdown_skill_dir>/scripts/run.sh" "<file_path>" -o "<workspace>/article.md"

Fallback: markdown-anything (cloud API, requires MDA_API_TOKEN)

MDA_API_TOKEN="<token>" bash "<markdown-anything_skill_dir>/scripts/convert.sh" "<file_path>" > "<workspace>/article.md"

After conversion: Read the Markdown output with encoding='utf-8'. Inspect the structure (headings, tables, images, code blocks). Use this to inform the HTML generation style.

Phase 3: Generate WeChat-Compatible HTML

Two approaches, pick based on content:

Approach A: Use md_to_wechat_html.py (quick, for structured documents) — PREFERRED

python "<skill_dir>/scripts/md_to_wechat_html.py" \
  --input "<workspace>/article.md" \
  --output "<workspace>/wechat_article.html" \
  --title "<article_title>"

This script converts Markdown directly to WeChat-ready inline-style HTML with Clockless design tokens. No additional CSS inlining step needed.

Approach B: Use html-anything (richer design, for complex content)

1. Load the html-anything skill.
2. Feed it the Markdown content as the source.
3. Follow html-anything workflow with WeChat compatibility rules (inline styles, light background, no CSS vars, max-width 680px, system fonts).
4. Then run anything-to-wechat's convert script:

python "<anything-to-wechat_skill_dir>/scripts/convert_for_wechat.py" \
  --input "<workspace>/wechat_article.html" \
  --output "<workspace>/wechat_article_final.html"

Decision guide: - PPTX, XLSX, data-heavy → Approach A (structured, table-friendly) - Long essays, research papers, narrative → Approach B (richer design) - Images, audio transcripts → Approach A (simpler layout)

Phase 4: Generate Cover Image

Use the ImageGen tool with a prompt derived from the article's topic and content.

• Save as wechat_cover.png in the workspace.
• Size: 1024x768 (WeChat cover ratio 4:3).
• Make it visually compelling.

Phase 5: Publish to WeChat Draft Box

IMPORTANT: On Windows, use Python subprocess to pass environment variables. The set VAR=val && python ... pattern does NOT work reliably on Windows CMD.

Cross-platform approach (recommended):

python -c "
import os, subprocess, sys
os.environ['WECHAT_APP_ID'] = '<app_id>'
os.environ['WECHAT_APP_SECRET'] = '<app_secret>'
result = subprocess.run([
    sys.executable,
    r'<anything-to-wechat_skill_dir>/scripts/publish_to_wechat.py',
    '--file', r'<workspace>/wechat_article.html',
    '--title', '<article_title>',
    '--cover', r'<workspace>/wechat_cover.png',
    '--digest', '<article_summary_under_120_chars>'
], capture_output=True, text=True, encoding='utf-8')
print(result.stdout)
print(result.stderr)
"

Or with environment variables already set:

python "<anything-to-wechat_skill_dir>/scripts/publish_to_wechat.py" \
  --file "<workspace>/wechat_article.html" \
  --title "<article_title>" \
  --cover "<workspace>/wechat_cover.png" \
  --digest "<article_summary_under_120_chars>"

Credentials: The publish_to_wechat.py script reads from WECHAT_APP_ID / WECHAT_APP_SECRET env vars. If not set, it will prompt interactively — the user just needs to paste their AppID and AppSecret.

Phase 6: Confirm & Handoff

Report success with Media ID and link to https://mp.weixin.qq.com/.

Tell the user: "文章已傳送到你的微信公眾號草稿箱，請登入微信公眾平臺稽核後一鍵釋出。"

WeChat HTML Compatibility

Same rules as anything-to-wechat:

• Inline styles only (no <style> tags)
• Light background (#ffffff)
• No CSS variables
• No position: fixed/sticky
• System fonts
• max-width: 680px
• Under 2MB total

Error Handling

Supported Input Formats

Script Reference

Configuration