### Start Web Server

Source: https://github.com/sansan0/trendradar/blob/master/README-EN.md

Start the web server for accessing generated reports.

```bash
docker exec -it trendradar python manage.py start_webserver
```

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

### Combined Configuration Example

Source: https://github.com/sansan0/trendradar/blob/master/README-EN.md

An example showing how to set both sorting priority and global display count limits in the configuration file.

```yaml
# config.yaml
report:
  sort_by_position_first: true   # Config order priority
  max_news_per_keyword: 10       # Global default max 10 per keyword
```

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

### Install Project Dependencies

Source: https://github.com/sansan0/trendradar/blob/master/README.md

Run the appropriate script to install project dependencies based on your operating system.

```bash
# Windows: setup-windows.bat or setup-windows-en.bat
# Mac/Linux: ./setup-mac.sh
```

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

### Build and Start All Services

Source: https://github.com/sansan0/trendradar/blob/master/README-EN.md

Build all TrendRadar services and start them in detached mode.

```bash
docker compose build
docker compose up -d
```

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

### Regex Start of String Matching

Source: https://github.com/sansan0/trendradar/blob/master/README-EN.md

Example of using regex to match titles that start with a specific pattern (`^pattern`).

```txt
/^breaking/
```

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

### Build and Start News Push Service

Source: https://github.com/sansan0/trendradar/blob/master/README-EN.md

Build and start only the news push service (trendradar).

```bash
docker compose build trendradar
docker compose up -d trendradar
```

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

### Clone Project and Start Services

Source: https://github.com/sansan0/trendradar/blob/master/README-EN.md

Clone the project repository and use docker compose to start both the news push and MCP services. Ensure you are in the 'docker' directory after cloning.

```bash
# Clone project (Recommended)
git clone https://github.com/sansan0/TrendRadar.git
cd TrendRadar/docker
docker compose up -d

# Check running status
docker ps | grep trendradar
```

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

### Clone Project and Navigate Directory

Source: https://github.com/sansan0/trendradar/blob/master/README-EN.md

Clone the TrendRadar repository and change into the project directory to begin setup.

```bash
git clone https://github.com/sansan0/TrendRadar.git
cd TrendRadar
```

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

### Start MCP Inspector

Source: https://github.com/sansan0/trendradar/blob/master/README-EN.md

Run this command to start the MCP Inspector, a debug tool for testing MCP connections. It is recommended to run this after starting the TrendRadar HTTP service.

```bash
npx @modelcontextprotocol/inspector
```

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

### LiteLLM Model Format Examples

Source: https://github.com/sansan0/trendradar/blob/master/README-EN.md

Examples of how to specify AI models using the 'provider/model_name' format with LiteLLM integration. This replaces the older 'provider' field and simplifies configuration.

```text
model: "openai/gpt-4o"
# DeepSeek: model: "deepseek/deepseek-chat"
# Gemini: model: "gemini/gemini-2.5-flash"
# Anthropic: model: "anthropic/claude-3-5-sonnet"

```

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

### Keyword Grouping Example

Source: https://github.com/sansan0/trendradar/blob/master/README-EN.md

Demonstrates how to define groups of keywords with required and filter words. Empty lines separate distinct groups.

```txt
iPhone
Huawei
OPPO
+launch

A-shares
Shanghai Index
Shenzhen Index
+fluctuation
!prediction

World Cup
Euro Cup
Asian Cup
+match
```

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

### Start TrendRadar HTTP Service

Source: https://github.com/sansan0/trendradar/blob/master/README-EN.md

Use these commands to start the TrendRadar HTTP service for MCP integration. Ensure you are in the project directory.

```bash
# Windows
start-http.bat
```

```bash
# Mac/Linux
./start-http.sh
```

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

### Global Filter Example

Source: https://github.com/sansan0/trendradar/blob/master/README-EN.md

Example demonstrating the effect of global filters on news content. Global filter words are excluded regardless of other matches.

```txt
[GLOBAL_FILTER]
advertisement

[WORD_GROUPS]
technology
AI
```

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

### Run HTTP Server with Custom Port

Source: https://github.com/sansan0/trendradar/blob/master/README.md

Start the MCP server using HTTP transport on a custom port, useful for troubleshooting port conflicts.

```bash
uv run python -m mcp_server.server --transport http --port 33333
```

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

### Run HTTP Server

Source: https://github.com/sansan0/trendradar/blob/master/README.md

Start the MCP server using HTTP transport on the default port 3333. This is used for viewing detailed error logs.

```bash
uv run python -m mcp_server.server --transport http --port 3333
```

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

### Start MCP Service Separately (Linux/Mac)

Source: https://github.com/sansan0/trendradar/blob/master/README-EN.md

Deploy the MCP service as a standalone container on Linux or macOS. Mounts the local 'config' and 'output' directories as read-only volumes and sets the timezone.

```bash
# Linux/Mac
docker run -d --name trendradar-mcp \
  -p 127.0.0.1:3333:3333 \
  -v $(pwd)/config:/app/config:ro \
  -v $(pwd)/output:/app/output:ro \
  -e TZ=Asia/Shanghai \
  wantcat/trendradar-mcp:latest
```

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

### Start Trendradar MCP Service

Source: https://github.com/sansan0/trendradar/blob/master/README-EN.md

Use this command to pull the latest image and start only the `trendradar-mcp` service for AI dialogue analysis.

```bash
docker compose pull trendradar-mcp
docker compose up -d trendradar-mcp
```

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

### Build and Start MCP AI Analysis Service

Source: https://github.com/sansan0/trendradar/blob/master/README-EN.md

Build and start only the MCP AI analysis service (trendradar-mcp).

```bash
docker compose build trendradar-mcp
docker compose up -d trendradar-mcp
```

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

### Required Word Syntax Example

Source: https://github.com/sansan0/trendradar/blob/master/README-EN.md

Demonstrates the syntax for specifying required words in the frequency_words.txt file. Both 'Tang Monk' and 'Pig' must appear in the title for the news to be included.

```text
+Tang Monk
+Pig
```

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

### Start All TrendRadar Services with Docker Compose

Source: https://github.com/sansan0/trendradar/blob/master/README-EN.md

Pull the latest Docker images and start all TrendRadar services, including push and AI analysis, in detached mode.

```bash
docker compose pull
docker compose up -d
```

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

### Regex Multiple Options Matching

Source: https://github.com/sansan0/trendradar/blob/master/README-EN.md

Example of using regex to match any one of several options using the pipe symbol (`|`). Note the escaped pipe `\|`.

```txt
/apple\|huawei\|xiaomi/
```

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

### Start MCP Service Separately (Windows PowerShell)

Source: https://github.com/sansan0/trendradar/blob/master/README-EN.md

Deploy the MCP service as a standalone container on Windows using PowerShell. Mounts the current directory's 'config' and 'output' folders as read-only volumes and sets the timezone.

```powershell
# Windows PowerShell
docker run -d --name trendradar-mcp `
  -p 127.0.0.1:3333:3333 `
  -v ${PWD}/config:/app/config:ro `
  -v ${PWD}/output:/app/output:ro `
  -e TZ=Asia/Shanghai `
  wantcat/trendradar-mcp:latest
```

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

### Start Only TrendRadar News Push Service with Docker Compose

Source: https://github.com/sansan0/trendradar/blob/master/README-EN.md

Pull the latest image for the TrendRadar service and start only the news push functionality in detached mode.

```bash
docker compose pull trendradar
docker compose up -d trendradar
```

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

### Configure Timezone Settings

Source: https://github.com/sansan0/trendradar/blob/master/README-EN.md

Set the application's timezone to ensure correct push times. Examples for Los Angeles and London are provided.

```yaml
app:
  timezone: "Asia/Shanghai"  # Default is China Standard Time
```

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

### Slack Message Format Preview

Source: https://github.com/sansan0/trendradar/blob/master/README-EN.md

This is an example of the message format sent to Slack. It includes batch information and trending topic details.

```text
*[Batch 1/2]*

📊 *Trending Topics Statistics*

🔥 *[1/3] AI ChatGPT* : 2 articles

  1. [Baidu Hot] 🆕 ChatGPT-5 Official Release *[1]* - 09:15 (1 time)

  2. [Toutiao] AI Chip Stocks Surge *[3]* - [08:30 ~ 10:45] (3 times)
```

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

### Custom Message Template

Source: https://github.com/sansan0/trendradar/blob/master/README-EN.md

An example of a custom JSON template that includes a 'text' field for content and a 'username' field.

```json
{"text": "{content}", "username": "TrendRadar"}
```

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

### ntfy Push Notification - Free Usage Test

Source: https://github.com/sansan0/trendradar/blob/master/README-EN.md

Test your ntfy free usage setup by sending a message to your subscribed topic. Ensure the topic name is hard-to-guess.

```bash
curl -d "Test message" ntfy.sh/your-topic-name
```

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

### Configure Region Display Order

Source: https://github.com/sansan0/trendradar/blob/master/README-EN.md

Adjust the order of sections in push messages by modifying the `display.region_order` list. This example places 'AI Analysis' at the top.

```yaml
display:
  region_order:
    - ai_analysis                  # Move to first line
    - new_items
    - hotlist
    - rss
    - standalone
```

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

### Configuration Strategy: Broad to Strict

Source: https://github.com/sansan0/trendradar/blob/master/README-EN.md

Illustrates a step-by-step approach to refining keyword configurations from broad terms to more specific ones using required and filter words.

```txt
# Step 1: Start with broad keywords for testing
Artificial Intelligence
AI
ChatGPT

# Step 2: After finding mismatches, add required words
Artificial Intelligence
AI
ChatGPT
+technology

# Step 3: After finding noise, add filter words
Artificial Intelligence
AI
ChatGPT
+technology
!advertisement
!training
```

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

### Configuration Strategy: Avoid Over-Complexity

Source: https://github.com/sansan0/trendradar/blob/master/README-EN.md

Shows a recommended approach to split complex keyword groups into smaller, more precise groups for better management.

```txt
Huawei
OPPO
Apple
Samsung
vivo
OnePlus
Meizu
+phone
+launch
+sales
!fake
!repair
!second-hand
```

```txt
Huawei
OPPO
+new product

Apple
Samsung
+launch

phone
sales
+market
```

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

### Enable and Configure Monitoring Platforms

Source: https://github.com/sansan0/trendradar/blob/master/README-EN.md

Enable trending platform crawling and specify sources in the `config/config.yaml` file. Suggest choosing 10-15 core platforms to avoid information overload.

```yaml
platforms:
  enabled: true                       # Enable trending platform crawling
  sources:
    - id: "toutiao"
      name: "Toutiao"
    - id: "baidu"
      name: "Baidu Hot Search"
    - id: "wallstreetcn-hot"
      name: "Wallstreetcn"
    # Add more platforms...
```

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

### Keyword vs. Platform Display Mode Comparison

Source: https://github.com/sansan0/trendradar/blob/master/README-EN.md

Illustrates the difference in output formatting between 'keyword' mode and 'platform' mode for push messages. 'Keyword' mode groups by topic, while 'platform' mode groups by source.

```text
# keyword mode (group by keyword)
📊 Trending Keywords Stats
🔥 [1/3] AI : 12 items
  1. [Weibo] OpenAI releases GPT-5 #1-#3 - 08:30 (5 times)
  2. [Zhihu] How to view AI replacing programmers #2 - 09:15 (3 times)

# platform mode (group by platform)
📊 Trending News Stats
🔥 [1/4] Weibo : 12 items
  1. [AI] OpenAI releases GPT-5 #1-#3 - 08:30 (5 times)
  2. [Trump] Trump announces major policy #2 - 09:15 (3 times)
```

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

### Clone Project and Configure

Source: https://github.com/sansan0/trendradar/blob/master/README-EN.md

Clone the TrendRadar repository and modify configuration files before building the Docker image.

```bash
git clone https://github.com/sansan0/TrendRadar.git
cd TrendRadar

# Modify config files
vim config/config.yaml
vim config/frequency_words.txt

# Use build version docker compose
cd docker
cp docker-compose-build.yml docker-compose.yml
```

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

### Regex End of String Matching

Source: https://github.com/sansan0/trendradar/blob/master/README-EN.md

Example of using regex to match titles that end with a specific pattern (`pattern$`).

```txt
/release$/
```

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

### Regex Word Boundary Matching

Source: https://github.com/sansan0/trendradar/blob/master/README-EN.md

Example of using regex with word boundaries (`\bword\b`) to match standalone words.

```txt
/\bword\b/
```

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

### Manually Update Images and Restart

Source: https://github.com/sansan0/trendradar/blob/master/README-EN.md

Manually pull the latest Crawler and MCP images, then restart the services.

```bash
docker pull wantcat/trendradar:latest
docker pull wantcat/trendradar-mcp:latest
docker compose down
docker compose up -d
```

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

### Docker Deployment for Bark Server

Source: https://github.com/sansan0/trendradar/blob/master/README-EN.md

Use this Docker command to deploy a self-hosted Bark server. Ensure ports are correctly mapped.

```bash
docker run -d \
  --name bark-server \
  -p 8080:8080 \
  finab/bark-server
```

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

### View Help Information

Source: https://github.com/sansan0/trendradar/blob/master/README-EN.md

Display help information for TrendRadar management commands.

```bash
docker exec -it trendradar python manage.py help
```

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

### Regex Non-Letter Boundary Matching

Source: https://github.com/sansan0/trendradar/blob/master/README-EN.md

Example of using regex with non-letter boundaries (`(?<![a-z])ai(?![a-z])`) for looser matching, suitable for mixed languages.

```txt
/(?<![a-z])ai(?![a-z])/
```

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

### Combined Regex, Normal Keywords, and Filter

Source: https://github.com/sansan0/trendradar/blob/master/README-EN.md

Shows how to combine regex for precise matching with normal keywords and a filter word (`!`) in a single configuration.

```txt
/\bai\b/
artificial intelligence
machine learning
!advertisement
```

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

### Configure Standalone Display

Source: https://github.com/sansan0/trendradar/blob/master/README.md

Set up the standalone display for specific platforms and RSS feeds to show their complete content in push notifications. Also configures the maximum number of items to display.

```yaml
display:
  regions:
    standalone: true                  # 推送中展示独立展示区（关闭不影响 AI 分析）

  standalone:
    platforms: ["zhihu", "weibo"]     # 这些平台的热榜给我完整显示
    rss_feeds: ["hacker-news"]        # 这些RSS源的内容给我完整显示
    max_items: 20                     # 最多显示多少条
```

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

### Enter Container for Debugging

Source: https://github.com/sansan0/trendradar/blob/master/README-EN.md

Gain interactive shell access to a running container for debugging purposes.

```bash
docker exec -it trendradar /bin/bash
```

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

### Build for ARM64 Architecture

Source: https://github.com/sansan0/trendradar/blob/master/README-EN.md

Set the DOCKER_ARCH environment variable to build images for ARM64 architecture.

```bash
export DOCKER_ARCH=arm64
docker compose build
```

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

### Unified Scheduling System Configuration

Source: https://github.com/sansan0/trendradar/blob/master/README-EN.md

Control crawling, pushing, and AI analysis timing with the timeline.yaml file. Supports preset templates like 'always_on', 'morning_evening', 'office_hours', 'night_owl', and 'custom'.

```yaml
schedule:
  preset: "always_on"
  # custom template example:
  # presets:
  #   my_custom_template:
  #     crawl:
  #       - "08:00-09:00"
  #     push:
  #       - "09:00-09:30"
  #     ai_analyze:
  #       - "09:30-10:00"

```

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

### Filter Word Priority Example

Source: https://github.com/sansan0/trendradar/blob/master/README-EN.md

Illustrates how filter words take precedence over required words. Even if 'Tang Monk' is a required word, news containing 'Tang Monk reciting sutras' will be filtered out.

```text
+Tang Monk
!Tang Monk reciting sutras
```

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

### Read Article Tool Usage

Source: https://github.com/sansan0/trendradar/blob/master/README-EN.md

Demonstrates the recommended workflow for reading article bodies using the new Jina AI Reader tools. First search for news, then read the article body via its URL.

```python
from jina.serve.runtimes.gateway.tool import search_news, read_article

# Search for news articles
results = search_news(query="keyword", include_url=True)

# Read the body of a specific article
article_body = read_article(url=results[0].url)

```

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

### Set Push Notification Schedule (Preset)

Source: https://github.com/sansan0/trendradar/blob/master/README.md

Enable the scheduling system and select a preset for push notification timing. The 'morning_evening' preset is recommended for a balance of real-time updates and daily summaries.

```yaml
schedule:
  enabled: true
  preset: "morning_evening"     # 改这里就行
```

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

### LiteLLM Configuration Changes

Source: https://github.com/sansan0/trendradar/blob/master/README-EN.md

Illustrates the configuration changes required for migrating to LiteLLM, including the removal of 'provider' and 'AI_PROVIDER' and the renaming of 'base_url' and 'AI_BASE_URL'.

```yaml
ai:
  # ai.provider is removed
  api_base: "http://localhost:11434/v1" # Renamed from ai.base_url

# Environment variables:
# AI_PROVIDER is removed
# AI_API_BASE is renamed from AI_BASE_URL

```

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

### Batch Read Articles Tool Usage

Source: https://github.com/sansan0/trendradar/blob/master/README-EN.md

Shows how to use the `read_articles_batch` tool to read multiple article bodies efficiently. This tool supports up to 5 articles and includes auto rate-limiting.

```python
from jina.serve.runtimes.gateway.tool import read_articles_batch

article_urls = ["url1", "url2", "url3"]
article_bodies = read_articles_batch(urls=article_urls)

```

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

### Real-Time Trending Hotness Algorithm Configuration

Source: https://github.com/sansan0/trendradar/blob/master/README-EN.md

Configuration for the 'Real-Time Trending Type' scenario, prioritizing ranking weight. Suitable for users who need to quickly understand current hot topics.

```yaml
advanced:
  weight:
    rank: 0.8
    frequency: 0.1
    hotness: 0.1
```

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

### Enable AI Analysis with Environment Variables

Source: https://github.com/sansan0/trendradar/blob/master/README-EN.md

Configure AI analysis by setting environment variables for the API key, model, and enabling the feature.

```env
AI_ANALYSIS_ENABLED=true
AI_API_KEY=sk-xxxxxx
AI_MODEL=deepseek/deepseek-chat
```

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

### Docker Environment Variables for Reporting

Source: https://github.com/sansan0/trendradar/blob/master/README-EN.md

Sets reporting behavior using environment variables, including sorting priority and maximum news per keyword.

```bash
SORT_BY_POSITION_FIRST=true
MAX_NEWS_PER_KEYWORD=10
```

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

### Keyword-Specific Display Count

Source: https://github.com/sansan0/trendradar/blob/master/README-EN.md

Demonstrates how to override the global display count limit for specific keywords using a special syntax in the frequency words file.

```txt
# frequency_words.txt
Tesla
Musk
@20              # Key focus, show 20 (override global)

Huawei           # Use global config, show 10

BYD
@5               # Limit to 5
```

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

### Display Current Configuration

Source: https://github.com/sansan0/trendradar/blob/master/README-EN.md

Show the current configuration settings of the TrendRadar service.

```bash
docker exec -it trendradar python manage.py config
```

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

### Configure MCP Server in Claude Desktop/Cline

Source: https://github.com/sansan0/trendradar/blob/master/README-EN.md

Configure the MCP server settings for Claude Desktop or Cline using a JSON configuration file. This specifies the URL and type of the MCP server.

```json
{
  "mcpServers": {
    "trendradar": {
      "url": "http://127.0.0.1:3333/mcp",
      "type": "streamableHttp"
    }
  }
}
```

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

### Verify Config Files Inside Container

Source: https://github.com/sansan0/trendradar/blob/master/README-EN.md

List the contents of the configuration directory inside the TrendRadar container.

```bash
docker exec -it trendradar ls -la /app/config/
```

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

### Configure AI Translation Settings

Source: https://github.com/sansan0/trendradar/blob/master/README-EN.md

Enable AI translation and specify the target language for translated content.

```yaml
ai_translation:
  enabled: true          # Enable translation
  language: "Chinese"    # Target language (Chinese, English, Japanese...)
```

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

### In-Depth Topic Hotness Algorithm Configuration

Source: https://github.com/sansan0/trendradar/blob/master/README-EN.md

Configuration for the 'In-Depth Topic Type' scenario, emphasizing frequency weight for sustained heat analysis. Ideal for investors and researchers needing deep trend analysis.

```yaml
advanced:
  weight:
    rank: 0.4
    frequency: 0.5
    hotness: 0.1
```

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

### GitHub Actions Workflow Configuration

Source: https://github.com/sansan0/trendradar/blob/master/README-EN.md

This is a link to the GitHub Actions workflow file for the crawler. It is mentioned as needing copy replacement in older versions.

```link
https://github.com/sansan0/TrendRadar/blob/master/.github/workflows/crawler.yml
```

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

### Manually Execute Crawler

Source: https://github.com/sansan0/trendradar/blob/master/README-EN.md

Manually trigger a single run of the news crawler.

```bash
docker exec -it trendradar python manage.py run
```

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

### ntfy Push Notification - Self-Hosting Docker Deploy

Source: https://github.com/sansan0/trendradar/blob/master/README-EN.md

Deploy ntfy for self-hosting using Docker. This command sets up the ntfy service, maps port 80, and persists cache data.

```bash
docker run -d \
  --name ntfy \
  -p 80:80 \
  -v /var/cache/ntfy:/var/cache/ntfy \
  binwiederhier/ntfy \
  serve --cache-file /var/cache/ntfy/cache.db
```

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

### View Container Logs

Source: https://github.com/sansan0/trendradar/blob/master/README-EN.md

Retrieve the last 100 lines of logs from a Docker container.

```bash
docker logs --tail 100 trendradar
```

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

### Configure VSCode (Cline/Continue) with TrendRadar STDIO Mode

Source: https://github.com/sansan0/trendradar/blob/master/README-EN.md

Configure Cline/Continue in VSCode to connect to TrendRadar's MCP service using STDIO. This is the recommended method and involves specifying the command and arguments.

```json
{
  "trendradar": {
    "command": "uv",
    "args": [
      "--directory",
      "/path/to/TrendRadar",
      "run",
      "python",
      "-m",
      "mcp_server.server"
    ],
    "type": "stdio",
    "disabled": false
  }
}
```

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

### Display Output Files

Source: https://github.com/sansan0/trendradar/blob/master/README-EN.md

List the output files generated by the TrendRadar service.

```bash
docker exec -it trendradar python manage.py files
```

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

### Configure Email Push in config.yaml

Source: https://github.com/sansan0/trendradar/blob/master/README-EN.md

Ensure HTML report generation is enabled in the configuration file for email push functionality to work correctly. If disabled, email push will fail with an error.

```yaml
storage:
  formats:
    sqlite: true
    txt: false
    html: true   # Must be enabled, otherwise email push will fail
```

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

### Basic Keyword Matching

Source: https://github.com/sansan0/trendradar/blob/master/README-EN.md

Capture news containing any of the specified normal keywords.

```txt
Huawei
OPPO
Apple
```

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

### Enable Incremental Push

Source: https://github.com/sansan0/trendradar/blob/master/README-EN.md

Shows how to enable incremental push notifications by setting the FOCUS_NEW_ONLY switch to True in main.py. This feature only sends notifications for new topics.

```python
FOCUS_NEW_ONLY = True
```

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

### Configure Email Password

Source: https://github.com/sansan0/trendradar/blob/master/README-EN.md

When configuring email notifications, use the generated authorization code for QQ Mail and 163/126 Mail, or an app-specific password for Gmail. Do not use your regular account password.

```text
EMAIL_PASSWORD fill authorization code, not QQ password
```

```text
EMAIL_PASSWORD fill app-specific password
```

```text
EMAIL_PASSWORD fill authorization code
```

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

### View Real-time Logs

Source: https://github.com/sansan0/trendradar/blob/master/README-EN.md

Display real-time logs from the TrendRadar service.

```bash
docker exec -it trendradar python manage.py logs
```

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

### Configure Multiple Notification Recipients

Source: https://github.com/sansan0/trendradar/blob/master/README-EN.md

In config/config.yaml, separate multiple notification URLs or tokens with a semicolon (;). This applies to platforms like Feishu, DingTalk, Telegram, and ntfy.

```bash
# Send to 3 Feishu groups
FEISHU_WEBHOOK_URL=https://hook1...;https://hook2...;https://hook3...
```

```bash
# Send to 2 DingTalk groups
DINGTALK_WEBHOOK_URL=https://oapi...;https://oapi...
```

```bash
# Send to 2 Telegram users (Match one-to-one)
TELEGRAM_BOT_TOKEN=tokenA;tokenB
TELEGRAM_CHAT_ID=userA;userB
```

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

### Normal Word with Display Name Alias

Source: https://github.com/sansan0/trendradar/blob/master/README-EN.md

Assign a custom display name to a normal keyword using the `=> alias` syntax.

```txt
deepseek => DeepSeek News
```

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

### TrendRadar STDIO Mode Client Configuration

Source: https://github.com/sansan0/trendradar/blob/master/README-EN.md

A basic configuration template for connecting a STDIO client to the TrendRadar MCP service. Replace `/path/to/TrendRadar` with the actual project path. This mode is recommended.

```json
{
  "name": "trendradar",
  "command": "uv",
  "args": [
    "--directory",
    "/path/to/TrendRadar",
    "run",
    "python",
    "-m",
    "mcp_server.server"
  ],
  "type": "stdio"
}
```

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

### Configure Push Content and Style

Source: https://github.com/sansan0/trendradar/blob/master/README-EN.md

Customize push message display and content by modifying the `report` and `display` sections in `config/config.yaml`. This includes setting display modes, ranking thresholds, and regional order.

```yaml
report:
  mode: "daily"                    # Push mode
  display_mode: "keyword"          # Display mode (v4.6.0 new)
  rank_threshold: 5                # Ranking highlight threshold
  sort_by_position_first: false    # Sorting priority
  max_news_per_keyword: 0          # Maximum display count per keyword

display:
  region_order:
    - new_items                    # New trending section
    - hotlist                      # Hotlist section
    - rss                          # RSS subscription section
    - standalone                   # Independent display section
    - ai_analysis                  # AI analysis section
```

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

### Global Display Count Limit Configuration

Source: https://github.com/sansan0/trendradar/blob/master/README-EN.md

Sets a maximum number of news items to display per keyword globally. A value of 0 means unlimited.

```yaml
report:
  max_news_per_keyword: 10  # Max 10 per keyword (0=unlimited)
```

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

### Global Filter Configuration

Source: https://github.com/sansan0/trendradar/blob/master/README-EN.md

Globally exclude content by listing keywords under the [GLOBAL_FILTER] region. This filter has the highest priority.

```txt
[GLOBAL_FILTER]
advertisement
promotion
marketing
shocking
clickbait

[WORD_GROUPS]
technology
AI

Huawei
HarmonyOS
!car
```

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

### AI Analysis Mode and Time Window Configuration

Source: https://github.com/sansan0/trendradar/blob/master/README-EN.md

Configure AI analysis behavior with independent control over the mode and time window. Supports custom execution periods and daily frequency limits.

```yaml
ai_analysis:
  mode: "follow_report" # Options: follow_report | daily | current | incremental
  time_window:
    execution_period: "08:00-18:00"
    daily_frequency_limit: 3

```

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

### Configure Cursor with TrendRadar STDIO Mode

Source: https://github.com/sansan0/trendradar/blob/master/README-EN.md

Configure Cursor to use TrendRadar's MCP service via STDIO. This is the recommended method and requires specifying the command to run the MCP server.

```json
{
  "mcpServers": {
    "trendradar": {
      "command": "uv",
      "args": [
        "--directory",
        "/path/to/TrendRadar",
        "run",
        "python",
        "-m",
        "mcp_server.server"
      ]
    }
  }
}
```