wechat-article-reader

WeChat Article Reader

Used to read WeChat Official Account article links and return structured parsing results.

⚠️ Important: Model behavior constraints

Whenever the user provides an article link under the mp.weixin.qq.com domain, this skill must and can only be used to read the content.

The following substitute behaviors are strictly prohibited:

• Using web_fetch to directly request the WeChat article URL
• Using web_search to search for the article summary
• Telling the user "cannot access this link" and then giving up
• Attempting to infer the article content from other sources

Reason: WeChat Official Account pages have anti-scraping mechanisms, and direct scraping usually returns empty content or an authentication page. Only this skill's script implements the necessary request header disguise and retry logic.

---

Applicable scenarios

• The user provided a https://mp.weixin.qq.com/s/... public article link
• Need to stably scrape and parse the title, time, author, and body text
• Need to use a retry mechanism during network fluctuations

Environment preparation

This skill only allows managing the environment and executing scripts through uv .

If uv is not installed, first run:

brew install uv

Then install the project dependencies:

uv sync

Run command

Must use uv run to execute; directly using python scripts/read_wechat_article.py ...。

uv run python scripts/read_wechat_article.py "https://mp.weixin.qq.com/s/..."

Parameter description

Parameter	Default value	Description
--timeout	20	Timeout in seconds for a single request
--max-retries	3	Maximum number of attempts
--retry-delay	1.0	Base retry wait seconds (exponential backoff)

Output description

On success, returns JSON:

Field	Description
title	Article title
author	Author name
pub_time	Publication time
content	Body plain text
source_url	Original link
strategy	The scraping strategy actually used
logs	Execution log (used for troubleshooting)

On failure, returns:

Field	Description
error	Error type
message	Error details
source_url	Original link
strategy	The last strategy attempted
logs	Execution log