菜单
Collect WeChat public account articles via RSSHub and publish to Feishu Docs. Includes RSS mirror discovery, deduplication, markdown generation, and automated cron scheduling.
默认使用会先检查来源的 Prompt；你也可以切换为直接命令，或下载本地副本。
决定是否安装前，请先阅读 SKILL.md，以及 SkillsMP 当前展示的配套文件。
用 Codex 或 Claude 帮你安装 复制这段 Prompt，粘贴到 Codex、Claude 或其他助手里，让它检查 Skill 页面并帮你完成安装。
直接命令不会经过审查 Prompt；运行前请先检查来源。
npx skills add https://github.com/tangzheng202202/hermes-skills --skill wechat-rss-collector命令会保持在同一行。复制前请横向滚动并检查完整内容。
想先保存到本地？可下载 SkillsMP 当前能够提供的文件。
Integrate Darwin evaluation system with skill-manager for automated skill quality assessment and evolution using AI optimization
DevOps tools - Docker, cron jobs, webhooks, troubleshooting
GitHub workflows - repos, PRs, issues, code review
正在显示 SKILL.md
基于 SOC 职业分类
| name | wechat-rss-collector |
| description | Collect WeChat public account articles via RSSHub and publish to Feishu Docs. Includes RSS mirror discovery, deduplication, markdown generation, and automated cron scheduling. |
| version | 2.0.0 |
| triggers | ["监控公众号","RSS抓取","飞书知识库","定时收集文章","wechat articles","公众号文章采集"] |
| prerequisites | ["RSSHub mirror availability","Feishu App credentials (APP_ID, APP_SECRET)","Feishu chat_id for notifications","Python 3.8+","feedparser, requests, beautifulsoup4"] |
Complete pipeline for monitoring WeChat public accounts via RSS and publishing to Feishu.
https://rsshub.app/wechat/mp/wx_zbus_xcbmtyn/{biz_or_id}https://rsshub.app/wechat/feeddd/{id} (for full text)https://rsshub.pseudoyu.com/wechat/sogou/{encoded_name} (may be deprecated)Fri, 27 Mar 2026 08:11:04 GMTGetting RSS URL for a WeChat Account:
__biz parameter from URL (base64 encoded)https://rsshub.app/wechat/mp/wx_zbus_xcbmtyn/{biz_value}1. Docs API (docx/v1/documents)
docs:doc:write permissionhttps://feishu.cn/docx/{document_id}2. Wiki API (wiki/v2/spaces/{space_id}/nodes)
wiki:wiki:write permissionspace_xxxxxxxx (from Wiki settings URL)3. IM Message API (im/v1/messages)
im:message:send, im:message.group_msg permissionsoc_) or Group (starts with oc_)Block Type Values (must be integers, not strings):
1 - Page (document root)2 - Text (standard paragraph)3 - Heading 14 - Heading 25 - Heading 36 - Heading 47 - Heading 58 - Heading 69 - Heading 710 - Heading 811 - Heading 912-46 - Other block types (list, code, quote, etc.)999 - CalloutExample: Adding text content to document:
def add_text_block(token: str, doc_id: str, parent_block_id: str, content: str):
url = f"https://open.feishu.cn/open-apis/docx/v1/documents/{doc_id}/blocks/{parent_block_id}/children"
headers = {"Authorization": f"Bearer {token}", "Content-Type": "application/json"}
data = {
"children": [
{
"block_type": 2, # Integer! Not "text"
"text": {
"elements": [
{
"type": "text_run",
"text_run": {
"content": content,
"text_style": {}
}
}
]
}
}
]
}
resp = requests.post(url, headers=headers, json=data)
return resp.json()
Common error: Using "block_type": "text" returns field validation failed with code 99992402. Must use integer values.
def get_tenant_access_token(app_id: str, app_secret: str) -> str:
url = "https://open.feishu.cn/open-apis/auth/v3/tenant_access_token/internal"
resp = requests.post(url, json={"app_id": app_id, "app_secret": app_secret})
return resp.json()["tenant_access_token"]
~/.wechat_articles/.history.jsonTelegram:
def send_telegram_message(bot_token: str, chat_id: str, message: str):
url = f"https://api.telegram.org/bot{bot_token}/sendMessage"
requests.post(url, json={"chat_id": chat_id, "text": message, "parse_mode": "HTML"})
Feishu:
def send_feishu_message(token: str, chat_id: str, message: str):
url = "https://open.feishu.cn/open-apis/im/v1/messages"
requests.post(url, headers={"Authorization": f"Bearer {token}"},
params={"receive_id_type": "chat_id"},
json={"receive_id": chat_id, "msg_type": "text",
"content": json.dumps({"text": message})})
# Normal run (today's articles only)
python3 ~/wechat_collector.py
# Test mode (3 articles per account)
python3 ~/wechat_collector.py --all
# Local only (no Feishu)
python3 ~/wechat_collector.py --no-feishu
# Specific date
python3 ~/wechat_collector.py --date 2026-04-11
/wechat/mp/wx_zbus_xcbmtyn/{id} - Returns "NotFoundError"/wechat/feeddd/{id} - Returns "NotFoundError"/wechat/sogou/{encoded_name} - May be deprecatedHISTORY_FILE = Path.home() / "wechat_articles" / ".history.json"
# Parse history file for today's articles when RSSHub fails
permission denied: wiki space permission denied, tenant needs edit permissionoc_, distinguish by contextfeeddd route for full content (when available)# Account RSS mapping (verified working as of 2026-04-11)
ACCOUNTS = {
"Draco正在VibeCoding": "https://rsshub.app/wechat/mp/wx_zbus_xcbmtyn/Draco_Daily",
"AI寒武纪": "https://rsshub.app/wechat/mp/wx_zbus_xcbmtyn/AI_himao",
"左岸AI": "https://rsshub.app/wechat/mp/wx_zbus_xcbmtyn/leftbankai",
"老金带你玩AI": "https://rsshub.app/wechat/mp/wx_zbus_xcbmtyn/wlaixkj"
}
# Notification targets
FEISHU_CHAT_ID = "oc_f9135e7516510373d91fed5756d62d7d" # P2P with Ningyao
TELEGRAM_CHAT_ID = "813280132" # Your Telegram chat ID
# Cron schedule
# 0 9 * * * /usr/bin/python3 /path/to/wechat_collector.py >> /var/log/wechat.log 2>&1
When web_extract returns "Unauthorized" for WeChat MP URLs (common as of 2026-04-11), use browser automation:
# Step 1: Navigate to article
browser_navigate(url="https://mp.weixin.qq.com/s/xxxxx")
# Step 2: Get full content as accessible tree
browser_snapshot(full=True)
# Returns parsed structure with headings, paragraphs, lists
# Step 3: Convert to markdown and save locally
# Title is in heading level 1
# Author is typically in text after "原创" or first text node
# Content follows paragraph structure
Key findings from 2026-04-11:
web_extract consistently fails with "Unauthorized: Invalid token"When RSSHub sources are unavailable, extract articles from local history:
from pathlib import Path
import json
def extract_from_history(target_date: str = None):
'''Extract today's articles from history file when RSS fails'''
from datetime import datetime
target_date = target_date or datetime.now().strftime("%Y-%m-%d")
history_file = Path.home() / "wechat_articles" / ".history.json"
if not history_file.exists():
return []
with open(history_file, 'r', encoding='utf-8') as f:
history = json.load(f)
articles = []
for link, info in history.items():
if info.get('date') == target_date:
articles.append({
'source': info.get('source'),
'title': info.get('title'),
'link': link,
'date': target_date
})
return articles
# Usage when RSSHub fails
articles = extract_from_history("2026-04-11")
for article in articles:
print(f"{article['source']}|{article['link']}")
When user provides a single WeChat article URL for immediate archiving:
Try web_extract first (fastest)
Browser fallback (when extract fails)
browser_navigate(url=article_url)
snapshot = browser_snapshot(full=True)
# Parse snapshot['snapshot'] for content tree
# Extract: title (heading level 1), author (first text node), date (emphasis/em), body (paragraphs)
Save locally
output_path = f"~/wechat_articles/YYYY-MM-DD_{source}_{title}.md"
# Write markdown with metadata header
Upload to Feishu
# Create document
doc = create_doc(token, title=f"[{source}] {title}")
doc_id = doc['data']['document']['document_id']
# Add content block (CRITICAL: use integer block_type, not string)
add_block(token, doc_id, block_id, content, block_type=2) # 2 = text
~/wechat_collector.py~/feishu_docs_uploader.py~/wechat_articles/YYYY-MM-DD/~/.wechat_articles/.history.json/var/log/wechat_collector.log (when using cron)