Xiaohui Skills

https://github.com/xiaohui-zhangxh/skills
Admin
A personal Cursor Agent Skills repository for sharing reusable skills across multiple projects,...
Tokens:6,157
Snippets:35
Trust Score:8
Update:6 months ago
Show doc for...
Tokens:
Raw
### Install Skill Repository (Bash)

Source: https://context7.com/xiaohui-zhangxh/skills/llms.txt

This section provides three methods for installing the skills repository: user-level symbolic linking (recommended for multi-project use), project-level usage, and direct copying into the project. It ensures skills are accessible by Cursor IDE.

```bash
# Method 1: User-level symbolic link (recommended, shared across projects)
git clone https://github.com/your-username/xiaohui-skills.git ~/xiaohui-skills
mkdir -p ~/.cursor/skills
ln -sfn ~/xiaohui-skills/skills/wechat-bot-notification ~/.cursor/skills/wechat-bot-notification

# Update skills
cd ~/xiaohui-skills && git pull

# Method 2: Project-level usage (current project only)
git clone https://github.com/your-username/xiaohui-skills.git .cursor/skills-repo
ln -sfn .cursor/skills-repo/skills/wechat-bot-notification .cursor/skills/wechat-bot-notification

# Method 3: Direct copy to project
cp -r .cursor/skills-repo/skills/wechat-bot-notification .cursor/skills/
```

--------------------------------

### Create New Skill Structure (Bash)

Source: https://context7.com/xiaohui-zhangxh/skills/llms.txt

This snippet demonstrates how to create the necessary directory structure and the mandatory SKILL.md file for a new AI agent skill. The SKILL.md file defines the skill's metadata, rules, and usage scenarios.

```bash
# Create skill directory structure
mkdir -p skills/my-new-skill

# Create SKILL.md file (required)
cat > skills/my-new-skill/SKILL.md << 'EOF'
---
name: my-new-skill
description: A brief description of the skill's purpose and trigger scenarios, for Agent auto-selection
license: MIT
compatibility: macOS, Linux
disable-model-invocation: false
---

# Skill Name

## Rule Description
A detailed description of the skill's execution rules and limitations.

## Applicable Scenarios
- Scenario 1: Describe when to use this skill
- Scenario 2: Describe another usage scenario

## Execution Steps
1. Step one
2. Step two
3. Step three
EOF

# Optional: Create scripts, resource directories
mkdir -p skills/my-new-skill/scripts
mkdir -p skills/my-new-skill/references
mkdir -p skills/my-new-skill/assets
```

--------------------------------

### Configure Environment Variables (Bash)

Source: https://context7.com/xiaohui-zhangxh/skills/llms.txt

This code illustrates how to configure environment variables required by skills, specifically the WECHAT_BOT_TOKEN for enterprise WeChat notifications. It shows methods for project-level, user-level, and direct environment variable export.

```bash
# Create configuration directory
mkdir -p .config/skills/envs

# Configure enterprise WeChat bot Token (project-level)
echo "your_webhook_token_here" > .config/skills/envs/WECHAT_BOT_TOKEN

# Or configure user-level Token (system-wide)
mkdir -p ~/.config/skills/envs
echo "your_webhook_token_here" > ~/.config/skills/envs/WECHAT_BOT_TOKEN

# Or use environment variable
export WECHAT_BOT_TOKEN="your_webhook_token_here"
```

--------------------------------

### 配置企业微信机器人Token

Source: https://github.com/xiaohui-zhangxh/skills/blob/main/skills/wechat-bot-notification/SKILL.md

提供了通过环境变量、项目配置文件或用户配置文件设置Token的命令行方法。

```bash
export WECHAT_BOT_TOKEN="your_webhook_token_here"
```

```bash
mkdir -p .config/skills/envs && echo "your_webhook_token_here" > .config/skills/envs/WECHAT_BOT_TOKEN
```

```bash
mkdir -p ~/.config/skills/envs && echo "your_webhook_token_here" > ~/.config/skills/envs/WECHAT_BOT_TOKEN
```

--------------------------------

### Read Token Configuration with Priority (Bash)

Source: https://context7.com/xiaohui-zhangxh/skills/llms.txt

This bash script defines a function `get_wechat_token` that reads the WECHAT_BOT_TOKEN with a specific priority: environment variable first, then project configuration file, and finally user configuration file. It returns the token or an empty string if not found.

```bash
#!/bin/bash
# Read WECHAT_BOT_TOKEN with priority

get_wechat_token() {
  # Priority 1: Environment variable
  if [ -n "$WECHAT_BOT_TOKEN" ]; then
    echo "$WECHAT_BOT_TOKEN"
    return 0
  fi

  # Priority 2: Project configuration file
  local project_config=".config/skills/envs/WECHAT_BOT_TOKEN"
  if [ -f "$project_config" ]; then
    token=$(head -n 1 "$project_config" | xargs)
    if [ -n "$token" ]; then
      echo "$token"
      return 0
    fi
  fi

  # Priority 3: User configuration file
  local user_config="$HOME/.config/skills/envs/WECHAT_BOT_TOKEN"
  if [ -f "$user_config" ]; then
    token=$(head -n 1 "$user_config" | xargs)
    if [ -n "$token" ]; then
      echo "$token"
      return 0
    fi
  fi

  # Token not found
  echo ""
  return 1
}

# Usage example
TOKEN=$(get_wechat_token)
if [ -n "$TOKEN" ]; then
  WEBHOOK_URL="https://qyapi.weixin.qq.com/cgi-bin/webhook/send?key=${TOKEN}"
  echo "Webhook URL: $WEBHOOK_URL"
else
  echo "Warning: WECHAT_BOT_TOKEN not configured, skipping send"
fi
```

--------------------------------

### 企业微信机器人消息请求数据格式

Source: https://github.com/xiaohui-zhangxh/skills/blob/main/skills/wechat-bot-notification/SKILL.md

展示了发送给企业微信机器人Webhook的JSON请求体结构，支持Markdown和文本两种格式。

```json
{
  "msgtype": "markdown",
  "markdown": {
    "content": "# 今日工作日报\n\n**日期**：2025-11-28\n\n..."
  }
}
```

```json
{
  "msgtype": "text",
  "text": {
    "content": "今日工作日报\n\n日期：2025-11-28\n\n...",
    "mentioned_list": ["@all"]
  }
}
```

--------------------------------

### POST /cgi-bin/webhook/send

Source: https://context7.com/xiaohui-zhangxh/skills/llms.txt

Sends a Markdown-formatted message to an Enterprise WeChat group using a webhook key.

```APIDOC
## POST /cgi-bin/webhook/send

### Description
Sends a message to an Enterprise WeChat group. The message can be formatted in Markdown or plain text. The webhook URL requires a valid key parameter.

### Method
POST

### Endpoint
https://qyapi.weixin.qq.com/cgi-bin/webhook/send?key={WECHAT_BOT_TOKEN}

### Parameters
#### Query Parameters
- **key** (string) - Required - The unique webhook token for the group bot.

#### Request Body
- **msgtype** (string) - Required - The type of message, either 'markdown' or 'text'.
- **markdown** (object) - Optional - Required if msgtype is 'markdown'. Contains the 'content' field with markdown text.
- **text** (object) - Optional - Required if msgtype is 'text'. Contains 'content' and optional 'mentioned_list'.

### Request Example
{
  "msgtype": "markdown",
  "markdown": {
    "content": "# Daily Report\n\n**Date**: 2025-11-28\n\n## Highlights\n- Completed Task A"
  }
}

### Response
#### Success Response (200)
- **errcode** (integer) - 0 indicates success.
- **errmsg** (string) - 'ok' indicates success.

#### Response Example
{
  "errcode": 0,
  "errmsg": "ok"
}
```

--------------------------------

### Send Enterprise WeChat Notification (Bash)

Source: https://context7.com/xiaohui-zhangxh/skills/llms.txt

This snippet shows how to send a Markdown-formatted daily report to an enterprise WeChat group using the Webhook API. It constructs the webhook URL using the WECHAT_BOT_TOKEN and sends a POST request with the message content.

```bash
# Send Markdown formatted daily report to enterprise WeChat group
# Token priority: Environment variable -> Project config -> User config
WEBHOOK_URL="https://qyapi.weixin.qq.com/cgi-bin/webhook/send?key=${WECHAT_BOT_TOKEN}"

curl -X POST "${WEBHOOK_URL}" \
  -H "Content-Type: application/json" \
  -d '{ 
    "msgtype": "markdown",
    "markdown": {
      "content": "# Today's Daily Report\n\n**Date**: 2025-11-28\n\n## Work Review\n- Completed Feature A development\n- Fixed Bug B\n- Code Review PR #123\n\n## Tomorrow's Plan\n- Continue developing Feature C\n- Write unit tests"
    }
  }'

# Expected response (success)
# {"errcode":0,"errmsg":"ok"}

# Send text message (alternative, supports @all)
curl -X POST "${WEBHOOK_URL}" \
  -H "Content-Type: application/json" \
  -d '{ 
    "msgtype": "text",
    "text": {
      "content": "Today's Daily Report\n\nDate: 2025-11-28\n\nCompleted: Feature A development",
      "mentioned_list": ["@all"]
    }
  }'
```

--------------------------------

### POST /cgi-bin/webhook/send

Source: https://github.com/xiaohui-zhangxh/skills/blob/main/skills/wechat-bot-notification/SKILL.md

Sends messages to an Enterprise WeChat group using a webhook. Supports Markdown and text message formats.

```APIDOC
## POST /cgi-bin/webhook/send

### Description
Sends messages to an Enterprise WeChat group via a webhook. This endpoint supports both Markdown and plain text message formats.

### Method
POST

### Endpoint
https://qyapi.weixin.qq.com/cgi-bin/webhook/send?key=YOUR_KEY

### Parameters
#### Path Parameters
- **key** (string) - Required - The unique key for your webhook.

#### Request Body
- **msgtype** (string) - Required - Specifies the message type. Can be 'markdown' or 'text'.
- **markdown** (object) - Required if msgtype is 'markdown' - Contains the Markdown message content.
  - **content** (string) - Required - The Markdown formatted message content.
- **text** (object) - Required if msgtype is 'text' - Contains the text message content.
  - **content** (string) - Required - The plain text message content.
  - **mentioned_list** (array) - Optional - A list of user IDs to mention (e.g., ["@all"]).
  - **mentioned_mobile_list** (array) - Optional - A list of mobile numbers to mention.

### Request Example
```json
{
  "msgtype": "markdown",
  "markdown": {
    "content": "# Daily Report\n\n**Status:** All systems operational.\n\n- Task 1 completed.\n- Task 2 in progress."
  }
}
```

### Response
#### Success Response (200)
- **errcode** (integer) - 0 for success.
- **errmsg** (string) - "ok" for success.

#### Response Example
```json
{
  "errcode": 0,
  "errmsg": "ok"
}
```

### Message Length Limits
- **Markdown messages**: Maximum 4096 characters.
- **Text messages**: Maximum 2048 characters.

If content exceeds limits, truncate or send in multiple messages.

### Sending Frequency Limits
- Each robot can send a maximum of 20 messages per minute.
- Exceeding the limit requires waiting or batch sending.
```

--------------------------------

### WeChat Bot Notification API

Source: https://github.com/xiaohui-zhangxh/skills/blob/main/skills/wechat-bot-notification/SKILL.md

This section details the process of sending notifications via the WeChat Bot, including token retrieval, message formatting, and the strict user-initiated sending protocol.

```APIDOC
## WeChat Bot Daily Notification Rules

### ⚠️ Strict Limitations

- **No Proactive Sending**: The WeChat Bot cannot proactively send notifications. It can only remind users that the notification feature is available.
- **User-Initiated**: Notifications are only sent when the user explicitly commands it (e.g., "Send daily report to WeChat").
- **Avoid Message Chaos**: Strict adherence to these rules prevents message clutter in WeChat groups.

### Token Configuration and Priority

The bot token is read with the following priority:

1.  **Environment Variable**: `WECHAT_BOT_TOKEN`
2.  **Project Configuration**: `.config/skills/envs/WECHAT_BOT_TOKEN` (in the workspace root)
3.  **User Configuration**: `~/.config/skills/envs/WECHAT_BOT_TOKEN`

### Token Configuration Methods

**Method 1: Environment Variable**
```bash
export WECHAT_BOT_TOKEN="your_webhook_token_here"
```

**Method 2: Project Configuration File** (Not committed to Git)
- Path: `.config/skills/envs/WECHAT_BOT_TOKEN`
- Content: Plain text, first line is the token (trimmed).
```bash
mkdir -p .config/skills/envs && echo "your_webhook_token_here" > .config/skills/envs/WECHAT_BOT_TOKEN
```

**Method 3: User Configuration File** (System-wide)
- Path: `~/.config/skills/envs/WECHAT_BOT_TOKEN`
- Content: Same as project configuration.
```bash
mkdir -p ~/.config/skills/envs && echo "your_webhook_token_here" > ~/.config/skills/envs/WECHAT_BOT_TOKEN
```

### Execution Flow

**Step 1: Get Token**
- The system attempts to retrieve the token based on the priority order (Environment Variable > Project Config > User Config).
- If a token is found, a Webhook URL is constructed: `https://qyapi.weixin.qq.com/cgi-bin/webhook/send?key={token}`.
- If no token is found, the sending step is skipped.

**Step 2: Format Daily Report Content**
- Extract today's report content.
- Format the content using Markdown (recommended) or plain text.

**Step 3: Remind User (Do Not Send Actively)**
- After generating and committing the report, remind the user: "You can use the WeChat notification feature to send the daily report."
- **Crucially, do not ask**: "Do you want to send today's report to WeChat?"
- **Wait for explicit user command**.

**Step 4: Send Message to WeChat Group (Only Upon User Command)**
- **This step is executed ONLY after the user explicitly commands it.**
- **Request Body Example (Markdown)**:
```json
{
  "msgtype": "markdown",
  "markdown": {
    "content": "# Today's Daily Report\n\n**Date**: 2025-11-28\n\n..."
  }
}
```
- **Request Body Example (Text)**:
```json
{
  "msgtype": "text",
  "text": {
    "content": "Today's Daily Report\n\nDate: 2025-11-28\n\n...",
    "mentioned_list": ["@all"] // Optional: @everyone
  }
}
```
- Send an HTTP POST request to the Webhook URL with `Content-Type: application/json`.

**Step 5: Error Handling**
- Configuration file not found: Skip sending, optionally log a warning.
- Incomplete configuration: Log a warning, skip sending.
- Network request failure: Log an error, do not interrupt the main report generation process.

### Key Rules Summary

- **Strictly No Proactive Sending**.
- **Remind users, do not ask**.
- **Wait for explicit user commands**.
- **Avoid message chaos**.
- **Optional feature**: Does not interrupt core functionality if unavailable or fails.
- **Error Handling**: All errors should be caught gracefully.
- **Message Format**: Fixed to Markdown.
- **Message Length**: Max 4096 characters; truncate or paginate if necessary.
- **Send Frequency**: Adhere to WeChat's rate limits (max 20 messages/min).
- **Token Priority**: Env Var > Project Config > User Config.
```

--------------------------------

### Define WeChat Work Markdown Message Payload

Source: https://github.com/xiaohui-zhangxh/skills/blob/main/skills/wechat-bot-notification/SKILL.md

JSON structure for sending formatted Markdown messages to a WeChat Work bot. It supports headers, bold text, and lists within the content field.

```json
{
  "msgtype": "markdown",
  "markdown": {
    "content": "# 标题\n\n**加粗**\n\n- 列表项1\n- 列表项2"
  }
}
```

--------------------------------

### Define WeChat Work Text Message Payload

Source: https://github.com/xiaohui-zhangxh/skills/blob/main/skills/wechat-bot-notification/SKILL.md

JSON structure for sending plain text messages to a WeChat Work bot, including support for mentions.

```json
{
  "msgtype": "text",
  "text": {
    "content": "消息内容",
    "mentioned_list": ["@all"],
    "mentioned_mobile_list": []
  }
}
```
Tokens:
Raw