TrendRadar: Open-Source AI Trend Monitor for Multi-Platform Aggregation and Smart Alerts

Without a tool like TrendRadar, monitoring public opinion and trends means manually checking multiple websites, apps, and RSS feeds, then filtering out noise. TrendRadar centralizes those sources and lets you filter by keywords or AI interests, so you only receive what matters. It also solves the distribution problem by pushing alerts to the chat tools you already use.

In an era of overwhelming fragmented information, TrendRadar shows how open-source software can combine self-hosted data aggregation with AI analysis and standards like MCP. It gives individuals and small teams a practical path to build their own trend and opinion radar without relying on closed platforms.

TrendRadar is licensed under GPL-3.0. You can use it commercially, even sell copies or offer paid support, but any distributed modifications must remain GPL-3.0. This makes it safe for internal business use but less ideal for embedding in a proprietary SaaS product. The default data source is the newsnow API, so long-term maintenance also means keeping an eye on that upstream project.

You don't have to write code. The Docker route is the most non-coder-friendly: pull the wantcat/trendradar image, launch the container, and open the built-in web server on port 8080. If you prefer a cloud-driven approach, use the GitHub Actions template by copying the repository, adding at least one notification webhook secret, and letting the job generate an HTML report that you can host on GitHub Pages or Cloudflare Pages. The only manual edits are in a YAML config file to choose platforms, keywords, and notification channels.

Start by setting the AI_API_KEY and AI_MODEL environment variables to enable AI analysis and translation. Then create an ai_interests.txt file with natural-language interests, such as 'I follow open-source AI model funding'. For deeper data conversations, run the MCP server (it listens on localhost:3333 by default) and connect it to MCP clients like Cursor, Cline, Continue, or Cherry Studio. After that, you can ask the AI client to summarize today's hot trends, compare sentiment across platforms, or answer a direct question like 'What's the biggest tech news today?' based on the aggregated trend data.

TrendRadar focuses on lightweight self-hosted aggregation and notifications. Deep AI features depend on external LLM APIs, and data completeness relies on your configured sources (e.g., the newsnow API) and storage retention. It does not provide a general-purpose web crawler for sites outside the supported hotlist and RSS sources. For arbitrary site scraping, you would need additional tooling.

# What is TrendRadar?

TrendRadar is a self-hosted, Python-based AI trend monitoring tool. It aggregates hot lists from 11+ platforms (Zhihu, Douyin, Bilibili, Weibo, and more) and supports RSS/Atom feed management. With keyword filtering, AI-driven analysis, and multi-channel push notifications, it turns fragmented public data into an actionable daily digest.

At the time of writing, the repository has over 61,000 GitHub stars and describes itself as an 'AI-driven public opinion & trend monitor with multi-platform aggregation, RSS, and smart alerts'.

Key Capabilities

• Aggregates hot lists from 11 default platforms: Zhihu, Douyin, Bilibili, WallstreetCN, Tieba, Baidu, CLS, The Paper, Phoenix, Toutiao, and Weibo.
• RSS/Atom subscription management with keyword grouping and freshness filters.
• Flexible keyword filtering: plain words, required words, block words, count limits, global filters, regex, and display aliases.
• AI interest-based filtering and translation: describe topics in natural language in ai_interests.txt; translate news into any language.
• AI analysis reports: trend overview, sentiment analysis, cross-platform insights.
• Hot trend tracking: rank changes, new item detection, persistence analysis.
• Multi-channel multi-account notifications: WeCom/WeChat, Feishu, DingTalk, Telegram, email, ntfy, Bark, Slack, generic webhook.
• MCP (Model Context Protocol) server: let AI clients like Cursor, Cline, Continue, and Cherry Studio chat with your trend data.
• HTML report with dark mode, search, tabs, export, and reading progress.

Deployment Options

• Docker (recommended): image wantcat/trendradar, optionally wantcat/trendradar-mcp.
• GitHub Actions: copy the template, configure secrets, and auto-generate a static report site on GitHub Pages or Cloudflare Pages. Note that free GitHub Actions requires a manual 'Check In' every 7 days.
• Local Python: install uv and run uv sync.

No GPU or model download is required. The default data source is the newsnow API; you can also self-host newsnow and set a custom api_url.

AI and MCP Integration

• Set AI_API_KEY and AI_MODEL environment variables to enable AI analysis, translation, and interest filtering.
• The optional MCP server listens on localhost:3333 and supports streaming output plus OpenAPI/Swagger docs.
• In ai_interests.txt, write something like 'I follow open-source AI model funding' to filter relevant news, or ask the system for a daily tech sentiment report.

Open Source and Commercial Use

TrendRadar is licensed under GPL-3.0. Commercial use is allowed, including selling copies or offering paid support, as long as you provide source code and license derivative works under GPL-3.0. Internal use typically does not trigger the copyleft requirement. The author asks for attribution and a link back to the repository when you use or reference the project.

Who Is It For?

• Investors/traders monitoring market and news trends.
• Content creators and self-media researchers gathering topic ideas.
• Corporate PR and brand managers tracking public opinion.
• General users who want a personalized news digest.
• Data analysts and AI enthusiasts who prefer chat-driven trend analysis via MCP.

Capability Ceiling

TrendRadar focuses on lightweight self-hosted aggregation and notifications. Deep AI features depend on external LLM APIs, and data completeness relies on your configured sources (e.g., the newsnow API) and storage retention. It does not provide a general-purpose web crawler for sites outside the supported hotlist and RSS sources. For arbitrary site scraping, you would need additional tooling.

Final Thoughts

TrendRadar combines aggregation, AI analysis, notifications, and MCP into one cohesive open-source project. If you want a personal public-opinion monitoring system up and running in under 30 minutes, this is one of the most straightforward options available.