Inoreader

Integrates inoreader to your claude desktop :D

python

8

GitHub Stars

python

Language

6 months ago

First Indexed

2 months ago

Catalog Refreshed

Documentation & install

Readme and setup notes from the catalogue, plus a client-ready config you can copy for your MCP host.

Installation

Add the following to your MCP client configuration file.

Configuration

View docs

{
  "mcpServers": {
    "dindicoelho-inoreader-mcp": {
      "command": "python",
      "args": [
        "/full/path/to/inoreader_mcp/main.py"
      ],
      "env": {
        "INOREADER_APP_ID": "YOUR_APP_ID",
        "INOREADER_APP_KEY": "YOUR_APP_KEY",
        "INOREADER_PASSWORD": "YOUR_PASSWORD",
        "INOREADER_USERNAME": "YOUR_EMAIL"
      }
    }
  }
}

This MCP server integrates Inoreader with Claude Desktop to help you manage, analyze, and derive insights from your RSS feeds. With it, you can list feeds and articles, read and mark items, search across your library, generate summaries, perform analytics, and view unread statistics all from natural language commands.

How to use

After you configure Claude Desktop to load the Inoreader MCP server, you can control your feeds using simple natural language requests. You can list all feeds, show articles with filters (such as unread status or by a specific feed), read full article content, and mark items as read either individually or in bulk.

You can perform searches across your subscribed feeds to find relevant articles. Generate summaries for individual articles or run analyses on multiple articles to obtain consolidated summaries, trend analysis, sentiment insight, and keyword extraction. You can also check statistics like the number of unread articles.

Typical usage patterns include: asking for the last unread articles, requesting summaries for the top AI articles of this week, asking for trends across your feeds today, or querying the overall sentiment of a topic like economics.

How to install

Prerequisites: you need Python installed on your system. You may also run the installer from Claude Desktop if you prefer a streamlined setup.

Auto-Installer (recommended): run the following command to install everything and configure Claude Desktop automatically.

python3 install_inoreader_mcp.py

Windows users can run the Windows installer by double-clicking the provided installer file.
Manual installation steps (if you prefer a hands-on approach): clone the repository, install dependencies, and set up credentials as shown.

git clone <repository-url>
cd inoreader_mcp
pip install -r requirements.txt
cp .env.example .env

Configure credentials inside the environment file and your Claude Desktop configuration to enable seamless operation.
Restart Claude Desktop after configuration so the MCP server is loaded and ready to accept commands.

Configuration and execution details

The MCP server is exposed to Claude Desktop as a local stdio-based server. You specify the command to run the server and the arguments needed to launch it, along with the required environment variables.

{
  "mcpServers": {
    "inoreader": {
      "command": "python",
      "args": ["/full/path/to/inoreader_mcp/main.py"],
      "env": {
        "INOREADER_APP_ID": "your_app_id",
        "INOREADER_APP_KEY": "your_app_key",
        "INOREADER_USERNAME": "your_email",
        "INOREADER_PASSWORD": "your_password"
      }
    }
  }
}

Troubleshooting and notes

If you encounter authentication errors, verify that your Inoreader credentials are correct and that the Inoreader application has the necessary permissions.

If the MCP server does not appear in Claude Desktop, double-check the full path to main.py in the configuration, restart Claude Desktop, and ensure Python is available in your system PATH.

For request timeouts, the Inoreader API may be slow. Consider reducing the number of articles requested in a single call.

Project structure and development notes

The MCP server is composed of a Python project with the following core components: the main MCP server entry, an Inoreader API client, MVP tool implementations, configuration management, and utility helpers. Development and testing can be done by running the server locally and inspecting console logs for debugging.

Available tools

list_feeds

List all subscribed feeds to show what you are following.

list_articles

List articles with filters such as unread status, by feed, or by time period.

read_article

Retrieve the full content of a specific article by ID.

mark_read

Mark articles as read either individually or in bulk.

search_articles

Search for keywords across your feeds.

summarize_articles

Generate summaries for individual articles or a set of articles.

analyze_articles

Perform consolidated summaries, trend analysis, sentiment analysis, and keyword extraction across multiple articles.

statistics

Provide unread article counters and feed statistics.