TrendRadar: Multi-Platform Trend Aggregator with AI Filtering and Push Alerts

What TrendRadar Solves: One Feed Instead of Many Tabs

Monitoring trending topics across multiple platforms means switching between them constantly, which produces noise as much as signal. TrendRadar addresses this by aggregating content from many sources into a single pipeline, applying keyword filters and AI-based relevance scoring, and delivering only the results that match a configured profile. The README describes its goal as lightweight and easy to deploy, and the project targets a reader who wants to stop scrolling through platforms and instead receive a filtered brief.

The tool supports RSS subscriptions alongside platform-specific hot-list sources, so it covers both real-time trending content and structured publication feeds. AI translation is also available, allowing monitoring of sources in one language while receiving summaries in another.

How TrendRadar Collects and Filters Data

TrendRadar fetches trending data from an external API provided by the newsnow project (github.com/ourongxing/newsnow). The README explicitly credits this dependency and asks users to control push frequency to avoid overloading the upstream service. This is a real architectural constraint: the data pipeline depends on a third-party API rather than direct platform scraping, so changes to that API can affect TrendRadar's data availability.

The pyproject.toml lists the key runtime dependencies: litellm 1.82.6 handles the AI model layer and supports multiple LLM providers through a unified interface, feedparser 6.0.12 processes RSS feeds, and json-repair 0.58.6 handles malformed JSON from API responses. Tenacity 0.5.0 adds retry logic. The project requires Python 3.12 or later.

After collection, the AI filtering step scores items against a configured topic profile. The README describes AI smart news filtering as a feature added in version 6.5.0. Items that pass the filter are packaged into a digest, optionally translated, and forwarded to the configured notification channels.

Deploying TrendRadar with Docker

The repository offers several deployment paths. Docker is the recommended route for most users, and the README describes it as a 30-second deployment. The Docker Hub images are published under the wantcat/trendradar and wantcat/trendradar-mcp names. The repository root also contains platform-specific setup scripts: setup-mac.sh for macOS and setup-windows.bat or setup-windows-en.bat for Windows, alongside start-http.sh and start-http.bat for launching the HTTP interface.

The pyproject.toml defines the project entry points:

[project.scripts]
trendradar = "trendradar.__main__:main"
trendradar-mcp = "mcp_server.server:run_server"

The project uses uv as its package manager (uv.lock is in the root), and requires Python 3.12 or later as stated in pyproject.toml. A Cloudflare deployment path is also mentioned in the README navigation, expanding the options beyond local or VPS hosting. Configuration lives in the config/ directory and controls sources, filters, notification channels, and AI provider credentials.

Notification Channels and AI Analysis Push

TrendRadar supports a wide range of push destinations. The README badges list WeChat Work (企业微信), personal WeChat, Telegram, Feishu (Lark), DingTalk, ntfy, Bark, Slack, and email. This breadth is useful for teams distributed across communication tools, but each channel requires its own API credentials and configuration.

The AI analysis push feature (added in version 5.0.0 according to the README navigation) sends AI-generated briefings directly to the configured channels rather than requiring the user to visit a dashboard. Version 5.2.0 added AI multi-language translation, allowing the digest language to differ from the source language. These features depend on a configured LLM provider accessed through litellm, so a valid API key for at least one supported provider is required for AI features to function.

MCP Integration for Conversational Analysis

A separate Docker image (wantcat/trendradar-mcp) exposes TrendRadar as an MCP server. MCP (Model Context Protocol) is a standard that allows AI coding assistants and chat interfaces to call external tools. With this integration, a user can ask questions in natural language inside an MCP-compatible client and have TrendRadar fetch and analyze trending data as part of that conversation.

The pyproject.toml entry point for this mode is:

trendradar-mcp = "mcp_server.server:run_server"

The repository includes README-MCP-FAQ.md and README-MCP-FAQ-EN.md, which document common questions about the MCP setup. The fastmcp 2.12.5 dependency handles the MCP server implementation.

Limitations Worth Knowing Before Deploying

The data pipeline depends on the newsnow API, which is a third-party service that the TrendRadar author does not control. The README explicitly asks users not to overwhelm this upstream source. If the newsnow service changes its API or becomes unavailable, the platform-specific hot-list data will stop flowing until TrendRadar is updated.

AI features require a valid LLM API key and incur API costs. The README mentions the project's own API costs as a reason for a small funding mechanism, which signals that the AI pipeline is not free at scale. The repository has no GitHub releases, so stable version tracking requires watching the repository directly or reading the version file. The current version listed in pyproject.toml is 6.10.0.

For users who want on-premises data with no third-party API dependency, TrendRadar's architecture does not currently support a fully self-contained data source for the platform hot-list feature.

TrendRadar vs. Standard RSS Readers

A conventional RSS reader such as FreshRSS handles feed aggregation and stores articles locally, but it does not apply AI scoring, generate summaries, or push filtered digests to external channels. It is also entirely feed-based and does not pull platform-specific hot lists.

TrendRadar sits in a different position: it is less a reading interface and more a pipeline. It collects, scores, and delivers without requiring the user to open a reader. The trade-off is that TrendRadar depends on an external data API for platform trends and requires an LLM API key for its AI features, whereas a standard RSS reader is self-contained and has no per-query cost. Teams that only need RSS aggregation will find TrendRadar heavier than necessary. Teams that want AI summarization and push delivery across multiple channels will find that a plain RSS reader does not cover those needs.

Editorial conclusion

TrendRadar fits individuals and teams that need a self-hosted, AI-augmented news digest without building one from scratch. The GPL-3.0 license requires that any distributed modifications be released under the same terms, which matters for teams considering commercial use. Before deploying, confirm that the notification channels you need are supported and that you have API access for whichever LLM provider you intend to use for AI filtering.

Frequently asked questions

Official sources

1. Official documentation
2. Official README
3. Project repository