RSS

You can run MCP RSS to automatically fetch, organize, and search through your RSS articles with a smart, token-conscious workflow. It supports both traditional feed management and advanced search features, including optional AI-powered semantic search, while keeping your reading lifecycle streamlined with a four-status model and daily digests.

How to use

Set up MCP RSS and connect it to your RSS feeds. You’ll browse article titles first to quickly skim what’s new, then fetch full content for any selections you want to read in depth. Use keyword or semantic search to discover articles across feeds, then organize items with the four statuses: unread, read, favorite, and archived. Retrieve today’s unread articles grouped by category with the daily digest and read full articles on demand using their IDs.

Workflow patterns you’ll use often:

How to install

Prerequisites you need before running MCP RSS.

• Node.js (v18+)
• Docker & Docker Compose (for PostgreSQL with pgvector)
• OpenAI API Key (optional, only for semantic search)

Follow these concrete steps to get started.

# Quick Start with Docker Compose

# 1) Install the package globally or clone the project
npm install -g mcp_rss
# OR for local development
git clone <repository-url>
c d mcp_rss
npm install

# 2) Start PostgreSQL with pgvector
docker-compose up -d

# 3) Configure environment variables
cp .env.example .env
# Edit .env with your settings

# 4) Build the project
npm run build

Additional setup and configuration

Create an environment file with the database and feed settings, plus optional OpenAI settings for embeddings.

# Example environment setup
DB_HOST=localhost
DB_PORT=5433
DB_USER=mcp_user
DB_PASSWORD=123456
DB_NAME=mcp_rss
OPML_FILE_PATH=./feeds.opml
RSS_UPDATE_INTERVAL=60
OPENAI_API_KEY=sk-your-key-here

Notes on running and testing

The system relies on PostgreSQL with the pgvector extension. You can run PostgreSQL via Docker Compose or manually as shown. If your OpenAI API key is not set, embeddings are skipped but all other features remain functional.

When you start the MCP server, you’ll interact with the server through your MCP client using the provided endpoints or local stdio setup.

Security and maintenance

Keep your OpenAI API key secure. Rotate credentials as needed and ensure your database credentials are stored in a secure environment. Regularly back up your PostgreSQL database and monitor logs for any embedding or feed-fetch issues.

Troubleshooting

If you encounter database connection issues, verify that PostgreSQL is running and that the configured host/port match your docker-compose settings. For OpenAI embedding problems, confirm API key validity and available credits; embeddings are optional and do not block normal operation.

Notes on performance and usage tips

Aim to set RSS_UPDATE_INTERVAL to a reasonable value for production. Use pagination when querying large result sets. Embeddings are stored as 1536-dim vectors, enabling fast semantic search with cosine similarity and allowing token-efficient browsing.

Summary of features you get

RSS feed management, advanced search, semantic search (optional), smart four-status workflow, daily digest, high-performance PostgreSQL with vector search, auto-deduplication, token-efficient reading, and scalable pagination for large feed collections.

Starting points for your setup

# Example JSON configuration for local stdio server (Claude Desktop style)
{
  "mcpServers": {
    "rss": {
      "command": "node",
      "args": ["/absolute/path/to/mcp_rss/dist/index.js"],
      "env": {
        "OPML_FILE_PATH": "/path/to/your/feeds.opml",
        "DB_HOST": "localhost",
        "DB_PORT": "5433",
        "DB_USER": "mcp_user",
        "DB_PASSWORD": "123456",
        "DB_NAME": "mcp_rss",
        "RSS_UPDATE_INTERVAL": "60",
        "OPENAI_API_KEY": "sk-your-key-here"
      }
    }
  }
}

Available tools

The MCP RSS server provides a set of tools to browse, search, and read articles. See the Tools section for detailed usage of each command.

Available tools

get_content

Get articles with basic filtering and pagination. Defaults to latest by pubDate, with token-efficient options to include excerpts or full content on demand.

search_articles

Search articles by keyword with optional category, date range, and status filters. Searches both title and content.

semantic_search

AI-powered semantic search using OpenAI embeddings to find conceptually similar articles. Requires OPENAI_API_KEY and articles from 2020 onward.

get_daily_digest

Retrieve today’s unread articles grouped by category, based on pubDate for today.

get_weekly_favorites

Return favorite articles from the last 7 days with titles only for quick weekly review.

get_article_full

Fetch full content for a specific article by its ID to read in depth.

get_sources

List RSS feed sources with pagination and optional filters like favorites or category.

set_tag

Update an article’s status to manage reading workflow (unread, read, favorite, archived).