Introduction

The body of WeChat Official Account articles is not available on the open web. To batch collect and archive them into your own knowledge base, you typically have to stitch together three layers yourself: the discovery logic for the Sogou WeChat entry point, body text parsing and format conversion, and the synchronization pipeline to IMA / Dify / Obsidian. Each layer requires handling rate limits, verification pages, deduplication, and incremental bookkeeping, which amounts to repetitive labor.

The Skill introduced below packages these three layers into a callable capability. It takes keywords or article URLs as input, outputs files in specified formats, and optionally syncs them to knowledge bases.

What is this

WeChat Official Account Content Scraper Skill, located at @user_9d5a2a39/wechat-article-scraper, published by user_9d5a2a39, sourced from the community, categorized under data-analysis, with the tag requires_api_key=true. It addresses the entire pipeline of WeChat Official Account articles, from discovery, scraping, and deduplication to multi-platform export.

Version information appears in two places in the materials: the SkillHub page field records 1.0.2, while the introduction text labels it v1.2.0. Both coexist; in actual use, the skill package installed locally is the standard.

On SkillHub, this skill has a rating of 100000 and 1288 downloads.

Core Functions

The following sections explain the capabilities listed in the materials.

Scraping Modes

Three modes covering different entry points:

  1. Sogou Search Discovery: Provide keywords to discover articles via the Sogou WeChat entry point.
  2. Direct URL Scraping: Provide an article link to directly scrape the body text.
  3. Batch Integration: Mixes the two methods above, suitable for batch archiving by topic.

Output Formats

Four body text formats: Markdown, Text, HTML, JSON. Additionally, a plain text Markdown table of contents is generated for easier subsequent retrieval.

Knowledge Base Synchronization

Three targets: IMA, Dify, and Obsidian, synchronized with one click. API Keys are read from environment variables preferentially; do not write keys into parameters or configuration files.

Deduplication

Three-fold judgment: canonical links, content hashes, and title similarity. When a duplicate is detected, the duplicate reason is returned rather than silently skipping, facilitating troubleshooting.

Filtering and Sorting

Supports three dimensions: date range, keywords, and relevance. Filtering occurs at the scraping stage to reduce invalid requests.

Offline Analysis

Post-scraping actions include: extractive summarization, high-frequency keyword extraction, and incremental bookkeeping of scraping history. Incremental bookkeeping means that upon re-running, only newly appearing articles are processed.

Network Layer

Rate limiting; retries only for 429, 5xx, and timeouts, while respecting the Retry-After header in responses; includes cache TTL; explicitly reports verification pages or rate-limited status instead of disguising them as successful scrapes.

Output Persistence

Defaults to writing to the system temporary directory. Export subdirectories must not traverse outside the vault or output root, and no binary files are retained within the skill package.

Installation and Activation

The scraped materials do not provide installation commands or specific configuration items, so this part is not speculated upon. To confirm the installation method, please check the official address below.

Two things are certain:

  1. The skill is marked requires_api_key=true. You need to prepare and configure the corresponding API Key before use; prioritize environment variables during configuration.
  2. It is based on the general SKILL.md format, so it can be mounted directly in Agent environments supporting this format.

Triggering

The skill comes with a set of trigger words. You can trigger it by stating your need directly without memorizing commands. Trigger words include:

WeChat Official Account, Official Account article, WeChat graphic message, scrape Official Account, WeChat scraping, Official Account search, export WeChat article, download WeChat article, Official Account content, WeChat knowledge base, Official Account batch, WeChat ZIP, IMA import, Dify import, Obsidian sync, WeChat full text, Official Account crawler, WeChat collection, graphic message collection, Sogou WeChat, WeChat article archive.

Typical Usage

The following three forms are derived from the capability descriptions in the official materials. Specific parameters should be based on the instructions within the skill package.

Single Article Scraping

Provide a single article URL to directly scrape the body text and output it as Markdown or Text for temporary citation.

Topic Discovery

Provide a set of keywords to follow the Sogou search discovery path, filtering by date range and keywords, and outputting a sorted list of articles plus body text. This is suitable for inventorying materials in a specific domain.

Batch Archiving

Combine discovery and scraping, with output landing in the temporary directory while performing incremental bookkeeping, followed by synchronization to IMA or Obsidian. Upon re-running, previously scraped articles are skipped via deduplication, processing only new additions.

Use Cases and Notes

Suitable for developers who need to continuously archive Official Account content and want the content to ultimately land in their own knowledge base, especially teams using Official Accounts as an information source for secondary organization.

Several points to note:

  1. Rate limiting is a hard constraint. The skill only retries for 429, 5xx, and timeouts. When encountering rate limiting, back off and do not bypass it.
  2. Verification pages are explicitly reported. If a verification page is returned instead of the body text, it means the entry point is blocked. In this case, manual intervention or switching entry points is required; do not repeatedly retry.
  3. Deduplication returns reasons. When a duplicate is hit, check the reason field to distinguish between false duplicates caused by link normalization and actual content duplication.
  4. Outputs default to the temporary directory. Content requiring long-term preservation must be moved to the vault or output root by yourself. Export path traversal will be blocked.
  5. Do not commit binary files. The skill package does not contain binaries. If export results contain images or other binary resources, you need to handle them yourself.
  6. API Keys go through environment variables. Do not write them into code or configuration files for submission.

Conclusion

The value of this Skill lies in eliminating the repetitive labor involved in the WeChat Official Account graphic message collection pipeline: discovery, scraping, deduplication, export, and synchronization are run in one go. Simultaneously, rate limiting and explicit status reporting are implemented at the network layer, ensuring failures are not disguised as successes.

Official Address: https://api.skillhub.cn/user_9d5a2a39/wechat-article-scraper
SkillHub Page: https://www.skillhub.cn/skills/user_9d5a2a39/wechat-article-scraper