# Feedly MCP > Feedly MCP lets you stop manually browsing news sites. Connect your Feedly account to any AI client and give your agent full control over how you track industry trends. Your agent can manage every subscription, pull the full text of articles, organize content into tags, and mark everything as read—all through natural conversation. ## Overview - **Category:** productivity - **Price:** Free - **Tags:** rss-aggregator, content-curation, industry-trends, news-tracking, research-boards, stream-intelligence ## Description You don't have time to scroll endless news feeds. This MCP lets your AI client take over your entire content curation process. Instead of opening dozens of tabs and hoping you remember which article was important, your agent acts like a dedicated research assistant that lives inside your workflow. It gives you natural language control over every feed—you can ask it to list all sources or pull the full text from specific articles for deep analysis. You can even manage your organization tags and mark entire reading lists as read with simple prompts. By connecting this MCP through Vinkius, you centralize your news aggregation; your AI becomes the single point of control for managing research boards and keeping up with what matters across every topic. ## Tools ### subscribe_to_feed This tool lets your agent start following a brand new news source. ### get_article_details It pulls the entire text and metadata for any specific article you mention. ### get_feed_metadata The agent fetches general information about a particular news stream or feed. ### get_user_profile You can retrieve basic details about your personal Feedly user account. ### get_stream_contents This tool retrieves a list of recent articles from an entire news stream or category. ### list_categories The agent lists all the custom categories you have set up in Feedly. ### list_subscriptions You can get a list of every feed source you are currently following. ### list_tags This tool shows all the personal tags you've used to categorize your content. ### mark_articles_as_read It marks one or more specified articles as read, clearing them from your unread count. ### unsubscribe_from_feed The agent stops following a specific news source you no longer care about. ## Prompt Examples **Prompt:** ``` List all my categories in Feedly. ``` **Response:** ``` I've retrieved your categories. You have 4 organized streams: 'Tech' (ID: user/.../category/Tech), 'AI & ML', 'Market News', and 'Design'. Which one should we check for new articles? ``` **Prompt:** ``` Show me the last 3 unread articles in the 'AI & ML' category. ``` **Response:** ``` Fetching unread articles from AI & ML... I've found 3 new entries, including 'New LLM Benchmarks' and 'The state of RAG'. Would you like the full text summaries for these? ``` **Prompt:** ``` Subscribe to this feed: 'https://example.com/rss' and add it to 'Tech'. ``` **Response:** ``` Done! I've successfully added the new feed to your Feedly account under the 'Tech' category. You'll now see articles from this source in your Tech stream. ``` ## Capabilities ### Manage subscriptions You can programmatically add new sources or stop following old ones. ### Check stream activity The agent retrieves the newest articles from any specific feed or category in real time. ### Fetch full article content It pulls complete text and all metadata for deep analysis by your AI client. ### Organize tags and categories You can list, manage, or update the custom tags and organizational categories you use in Feedly. ### Mark articles as read The system automatically marks one or more selected articles as consumed, keeping your reading list clean. ## Use Cases ### Quickly auditing research scope A market researcher needs to know exactly what feeds are contributing to a project. They ask their agent, 'List all my sources for Q3 planning.' The agent runs `list_subscriptions`, immediately providing an inventory list so they don't have to navigate multiple screens. ### Processing competitor news A content strategist receives a batch of articles. They prompt the agent: 'Pull the full text and metadata for these 5 links.' The agent executes `get_article_details` on each, allowing them to run deep analysis without leaving their main workspace. ### Clearing old reading lists After a big review session, the editor asks the agent to mark all articles from 'Tech' as read. The agent runs `mark_articles_as_read` for that category, instantly clearing clutter and focusing attention on truly new content. ### Onboarding a new topic A professional needs to follow a niche industry blog. They instruct the agent: 'Follow this RSS feed and add it under my Finance category.' The agent runs `subscribe_to_feed` and handles the entire setup. ## Benefits - Automate reading cleanup. Instead of manually marking articles as read, the agent runs `mark_articles_as_read` when you're done, keeping your unread count accurate and manageable. - Deep dive into sources. When a specific article catches your eye, use `get_article_details` to pull the complete text for summarization, avoiding the need to copy-paste content manually. - Maintain perfect structure. You can list all tags using `list_tags` or manage categories via `list_categories`, ensuring every piece of research is properly filed without manual effort. - Centralized monitoring. If you're tracking a major industry event, your agent checks the latest entries from multiple sources using `get_stream_contents`, giving you one comprehensive view. - ## How It Works The bottom line is that once connected, your AI client treats Feedly like another built-in app, responding to natural requests instead of requiring manual API calls. 1. First, subscribe to this MCP and retrieve your Developer Access Token from Feedly's integration settings. 2. Next, connect the token within any MCP-compatible client (like Cursor or Claude) so your agent can authenticate with the service. 3. Finally, tell your agent what you need—for example, 'Show me all unread articles in my Market News category,' and it executes the command. ## Frequently Asked Questions **How do I list all my Feedly categories using the Feedly MCP?** You use the `list_categories` tool. This immediately returns a structured list of every category you have set up, letting your agent know exactly what organizational streams exist. **Can I pull full text from an article using Feedly MCP?** Yes, the `get_article_details` tool fetches the complete body and metadata of any article. This is crucial when your agent needs to summarize or analyze content deeply. **Is Feedly MCP better than just using RSS feeds? ** Yes, because this MCP adds intelligence. It lets your agent interact with the feed—you can list sources (`list_subscriptions`) and manage them programmatically, which standard RSS readers can't do. **What happens when I use mark_articles_as_read?** Running `mark_articles_as_read` updates your reading status across Feedly. Your agent handles the necessary API calls to ensure those articles are marked as consumed. **Can my AI client subscribe to a new feed with Feedly MCP?** Absolutely. You can use `subscribe_to_feed` to add any new news source directly through your agent, keeping your research pipeline current.