TrendRadar

TrendRadar MCP Server provides a programmable analytics backend that exposes a Model Context Protocol (MCP) interface. It lets you drive AI analytics and news trend processing locally or on your own infrastructure, so you can build, test, and extend advanced news insights with your preferred MCP clients.

How to use

You access TrendRadar’s MCP server from compatible MCP clients. Use the HTTP endpoint to connect remotely, or run the local MCP service in stdio mode from your machine. Once connected, you can query historical and live news data, request topic trends, and drive AI-enabled analyses using MCP clients such as Claude Desktop, Cursor, or other MCP-enabled tools.

How to install

Prerequisites: you will need Docker for a quick, reliable setup, or a local runtime if you choose a manual install. The following steps show a Docker-based installation flow and an alternative local runtime flow.

# Option A: Docker Compose (recommended)
# 1. Clone the TrendRadar repository
git clone https://github.com/sansan0/TrendRadar.git
cd TrendRadar

# 2. Start all services (news push + MCP analysis)
# This pulls the latest images and runs the containers in background
docker compose pull
docker compose up -d

# 3. Verify services
docker ps | grep trend-radar

# Option B: Run MCP service separately (local runtime)
docker run -d --name trend-radar-mcp \
  -p 127.0.0.1:3333:3333 \
  -v ./config:/app/config:ro \
  -v ./output:/app/output:ro \
  -e TZ=Asia/Shanghai \
  wantcat/trendradar-mcp:latest

# Test MCP endpoint
curl http://127.0.0.1:3333/mcp

Additional sections

Configuration and runtime details help you tailor deployment to your needs. The MCP server supports an HTTP connection when you want to run MCP remotely, and a stdio pattern for local execution. If you run in Docker, you can override settings with environment variables or a .env file for easier management.

# Example MCP server connection (stdio mode)
# command: uv
# args: ["--directory","/path/to/TrendRadar","run","python","-m","mcp_server.server"]

Security and notes

MCP endpoints exposed locally are intended for trusted clients. When using remote connections, ensure you protect the endpoint with appropriate network access controls and authentication as needed. If you deploy in a shared environment, restrict access to the MCP port and consider using a gateway or reverse proxy for added security.

Troubleshooting and tips

If the MCP service does not start, verify the port 3333 is not in use, confirm dependencies are installed, and check service logs for errors. For Docker setups, ensure your .env or environment variables are correctly wired, and that the TrendRadar output directory is writable by the container.

Examples and usage notes

You can connect MCP clients to the TrendRadar MCP server in two ways. Use the HTTP transport to point to the local or remote endpoint, or run the local stdio transport with the MCP runner. When using multiple MCP clients, you can clear and reuse the same endpoint across clients or configure separate endpoints per client as needed.

Tools and endpoints (high level)

The MCP layer exposes utilities for querying news data, performing trend analyses, and managing system state. Key capabilities include retrieving latest news, analyzing topic trends, and obtaining summaries from historical data.

Available tools

get_latest_news

Fetches the most recent news items from configured data sources.

get_news_by_date

Retrieves news items within a specified date range.

get_trending_topics

Returns topics currently trending across monitored platforms.

search_news

Performs keyword-based searches across stored news data.

search_related_news_history

Finds related historical articles to a given topic.

analyze_topic_trend

Analyzes the trend of a topic over time.

analyze_data_insights

Provides data-driven insights from collected news data.

analyze_sentiment

Performs sentiment analysis on news items.

find_similar_news

Finds news items similar to a given article.

generate_summary_report

Generates concise summaries of gathered news data.

get_current_config

Returns the current MCP server configuration.

get_system_status

Reports the status of running MCP services.

trigger_crawl

Manually trigger a data crawl from configured sources.