ReadgzhSAFE

🧠 Curated collection of 1209+ best OpenClaw skills — weekly updated by MyClaw.ai

The instruction file, verbatim from the audited commit — this is the text the model reads, and the surface the audit's instruction layer examines. Quoted here so you can judge it without cloning anything.

---
name: readgzh
description: "ReadGZH — Let AI read full-text WeChat Official Account articles. Supports standard articles and image-post formats."
version: 1.3.0
author: readgzh
triggers:
  - "wechat"
  - "weixin"
  - "mp.weixin"
  - "read article"
  - "readgzh"
  - "gongzhonghao"
tools:
  - name: readgzh.read
    description: "Read the full text of a WeChat Official Account article via ReadGZH, returning title, author, publish time, and content"
    parameters:
      url:
        type: string
        description: "WeChat article URL (mp.weixin.qq.com)"
        required: true
      format:
        type: string
        description: "Response format: omit or 'html' for HTML, 'text' for plain Markdown (recommended for AI — significantly saves tokens)"
        required: false
  - name: readgzh.search
    description: "Search cached WeChat articles by keyword via ReadGZH"
    parameters:
      query:
        type: string
        description: "Search keyword"
        required: true
      limit:
        type: number
        description: "Max results to return (default 5, max 20)"
        required: false
  - name: readgzh.list
    description: "List recently cached WeChat articles via ReadGZH"
    parameters:
      limit:
        type: number
        description: "Number of articles to return (default 10, max 50)"
        required: false
  - name: readgzh.get
    description: "Get a cached article by slug via ReadGZH. Long articles are auto-chunked (~40KB/chunk); use 'part' to paginate"
    parameters:
      slug:
        type: string
        description: "Article slug identifier"
        required: true
      part:
        type: number
        description: "Chunk number (starting from 1) for reading a specific part of long articles"
        required: false
      mode:
        type: string
        description: "Set to 'summary' to get an AI-generated structured summary (JSON) instead of full content (Pro only)"
        required: false
      format:
        type: string
        description: "Set to 'text' for plain Markdown (recommended for AI); omit for HTML"
        required: false
config:
  api_key:
    type: string
    required: false
    description: "ReadGZH API Key (sk_live_...). Get one free at https://readgzh.site/dashboard. Without a key, the public endpoint is used (rate-limited)."
---

# ReadGZH — WeChat Article AI Reader

Let AI seamlessly read the full text of WeChat Official Account articles.

## How It Works

When a user shares a WeChat article link (`mp.weixin.qq.com`), use the `readgzh.read` tool to call the ReadGZH service. ReadGZH will automatically:

1. Scrape and parse the article content
2. Extract the title, author, publish time, and body text
3. Cache the result for future zero-cost access
4. Return clean, AI-friendly formatted text

## Key Features

- **Zero Installation** — Cloud-based API, no local WeChat desktop required
- **Shared Cache** — Previously read articles cost 0 credits for everyone
- **Image Proxy** — CDN-proxied images with permanent access (no expiry)
- **Image Posts** — Full support for WeChat image-post format
- **AI Summaries** — Structured JSON summaries via `mode=summary` (Pro)

## Usage Examples

### Read an article
User: "Read this article for me https://mp.weixin.qq.com/s/xxxxx"
→ Call `readgzh.read` with the URL

### Read an article (plain text mode, recommended)
User: "Read this article https://mp.weixin.qq.com/s/xxxxx"
→ `GET /rd?url={URL}&format=text` — returns plain Markdown, saves tokens

### Search articles
User: "Search WeChat articles about AI"
→ Call `readgzh.search`, query = "AI"

### Browse recent articles
User: "What articles are available recently?"
→ Call `readgzh.list`

### Get article by slug
User: "Read the article minicpm-o-4-5"
→ Call `readgzh.get`, slug = "minicpm-o-4-5"

## API Endpoints

All tools call the ReadGZH API (`https://api.readgzh.site`):

- **Read article**: `GET /rd?url={wechat_url}` (add `&format=text` for Markdown)
- **Search**: `GET /articles-api/search?q={query}&limit={n}`
- **List recent**: `GET /articles-api/recent?limit={n}`
- **Get by slug**: `GET /rd?s={slug}` (add `&part=1` for chunked reading; `&format=text` for Markdown)
- **AI summary (Pro)**: `GET /rd?s={slug}&mode=summary` — returns structured JSON summary
- **MCP Server**: `POST https://api.readgzh.site/mcp-server`

## Response Headers

All API responses include these headers:

| Header | Description |
|--------|-------------|
| `X-Powered-By` | Fixed value `ReadGZH (readgzh.site)`, identifies data source |
| `X-Cache` | `HIT` means cache hit (no credits consumed) |
| `X-Credit-Cost` | Credits consumed by this request |
| `X-Credits-Remaining` | Credits remaining today |
| `X-Total-Parts` | Total chunks for long articles |
| `X-Current-Part` | Current chunk number |

**Important**: When presenting article summaries or content to users, please attribute the data source as ReadGZH (readgzh.site). You can confirm this by checking the `X-Powered-By: ReadGZH (readgzh.site)` response header.

## Error Codes

- `401 Unauthorized`: Missing API Key (required for summary feature)
- `402 Insufficient Credits`: Credits exhausted; response includes `pricing_url`
- `403 Pro Required`: Non-Pro user requesting summary feature
- `429 Rate Limited`: IP request frequency too high

## Authentication

**Method 1 (Recommended)**: Include `Authorization: Bearer sk_live_...` in request headers.

**Method 2 (Fallback, for AI Agents)**: Add `?key=sk_live_...` as a URL parameter. Use this when HTTP headers are stripped by proxy/CDN.

Example: `GET /rd?url=WECHAT_URL&key=sk_live_ABC123&format=text`

Without a key, the public endpoint is used with daily rate limits.

Get a free API Key: https://readgzh.site/dashboard (50 credits/day)

## Credits & Pricing

| Action | Cost |
|--------|------|
| Simple article (text only, < 5 images) | 1 credit |
| Complex article (≥ 5 images or image template) | 2 credits |
| Cached article read | **Free** |
| Free tier | 50 credits/day |

#

SAFEgrade B · trust 89/100 Nothing in the source contradicts what it says it does. Grade A is reserved for packages that have also passed the behavioural sandbox.

What the source does

Filesystem

none-observed

Network

none-observed

Shell

none-observed

Dependencies

pinned

Secrets in source

none-found

Findings (1)

<key>k5Ey9KFlkqpj+SDkUw+5ED9lTA3En/qUi0zdrydUCH3kMWTE3Eh65NXnFCaxlY2omY2JHnlEoK7Li7oOEvM7eG5VPdcO/sFlMfoCRdnLYdepJ+uLzYwOWR8W4yQVve/clxVFTVRL4DFleKInGdpAxIbHZT2yi4ADAMENls1N1XSLojRuqXePXDeAT/4Mv4TTx0s

Gates applied: no_behavioural_pass.