trendradar

TrendRadar (sansan0/TrendRadar) is an open-source tool for monitoring trending topics and public sentiment, designed to be lightweight and easy to deploy, helping users escape information overload by focusing only on the news they truly care about. Implemented in Python, its core features include: aggregation of multi-platform hot lists (via newsnow API), RSS/Atom feed scraping, precise keyword filtering (supports regular expressions and group aliases), AI-powered filtering (users describe interests in natural language in ai_interests.txt; AI extracts tags and scores articles, falling back to keyword matching if anomalies occur), AI multilingual translation, AI-generated analytical briefings (trend summaries, keyword heat, sentiment analysis, cross-platform correlations), and MCP services enabling AI clients to query news, search, read articles, compare time periods, etc., using natural language. Notification channels include WeChat Work, personal WeChat, Feishu, DingTalk, Telegram, email, ntfy, Bark, Slack, and generic Webhooks. Deployment options support GitHub Actions + GitHub Pages, Docker (including standalone image for MCP service), local execution, and Cloudflare Pages; storage supports local SQLite and S3-compatible cloud storage (e.g., Cloudflare R2), allowing self-hosted data. Configuration includes a visual editor for config.yaml / timeline.yaml, preset time templates, custom scheduling, toggleable display regions, remote batch uploads. The project generates interactive HTML reports with dark mode, search, keyboard shortcuts, and Markdown export. Licensed under GPL-3.0. The README’s changelog indicates current version v6.10.0, with ongoing iteration of the MCP toolkit.