TrendRadar: Precision AI Alerts via MCP

TrendRadar is an AI-driven public opinion and trend monitor, aggregating hot topics from multiple platforms and RSS feeds to deliver smart, filtered alerts. With a significant 61,483 GitHub stars, this Python-based MCP Server is engineered to cut through information overload, providing developers with a structured way to integrate real-time trend data into their AI applications.

MCP-Driven Multi-Channel AI Delivery

The core of TrendRadar's MCP integration lies in its ability to push AI-generated content directly to various communication channels. As of mcp-v4.0.0 (released 2026/02/09), AI-written messages can be sent to 9 different platforms, including Feishu, DingTalk, Telegram, and email. The system automatically adapts Markdown formatting for each platform, eliminating manual adjustments.

A key feature for developers is the get_channel_format_guide tool. This exposes channel-specific formatting rules and limitations to an AI model, enabling it to generate better-laid-out content. For handling lengthy messages, TrendRadar intelligently splits content according to channel-specific byte limits (e.g., Feishu's 30KB, DingTalk's 20KB), with these configurations managed via config.yaml. The batch processing functions directly reuse TrendRadar's core modules, ensuring code efficiency.

Extracting and Analyzing Content with MCP Tools

TrendRadar exposes a suite of tools for content extraction and analysis, crucial for any AI system aiming to understand public discourse.

Key tools include:

• search_news(query="关键词", include_url=True): Initiates a search for news articles based on a query, with an option to include URLs.
• read_article(url=...): Fetches the full text of a single article in Markdown format using Jina AI Reader.
• read_articles_batch: Reads up to 5 articles in a batch, with automatic rate limiting.
• get_latest_rss: Retrieves the latest RSS entries, supporting multi-day queries and cross-date URL deduplication.
• search_rss: Searches through RSS feeds.
• get_rss_feeds_status: Provides the status of configured RSS feeds.
• aggregate_news: Performs cross-platform news deduplication and aggregation.
• compare_periods: Conducts period-over-period analysis (e.g., week-on-week, month-on-month).
• find_related_news: Combines previous functionalities to find similar and related news.
• get_trending_topics: Enhanced with an auto_extract mode to automatically identify trending topics and support /pattern/ regular expressions.

All 21 tool functions are wrapped with asyncio.to_thread() for asynchronous consistency, and their return values are standardized to a {success, summary, data, error} structure. TrendRadar also exposes MCP Resources for platforms, rss-feeds, available-dates, and keywords.

AI-Powered Filtering and Translation

TrendRadar's core AI capabilities extend to filtering and translation. The filter method can be set to ai in config.yaml, with a min_score threshold (1-10) for pushing content. This AI filtering shares model configurations with AI analysis and translation.

For multi-language support, the ai_translation feature can be enabled in config.yaml, allowing content to be translated into any language, such as English, Korean, or Japanese. This feature supports custom translation styles via ai_translation_prompt.txt and uses intelligent batch processing to optimize API calls.

# config.yaml quick start example for AI translation
ai_translation:
  enabled: true
  language: "English"  # Target translation language

This makes TrendRadar a strong candidate for developers building global AI-driven news analysis or sentiment platforms, allowing them to process and disseminate information across linguistic barriers and diverse communication channels efficiently.

References

• TrendRadar on GitHub
• Model Context Protocol Documentation
• TrendRadar on model-context-protocol.com