Check and validate RSS/Atom feeds, find a site's feeds, read latest items, search 1,600+ feed sites
Maintain this server? Claim it
Call this server's tools from your browser. Nothing to install.
Requests go through mcptry because browsers can't call most MCP servers directly. Header values are sent to the server for that request only and are never stored or logged.
Tool list from our last check. Connect live to refresh it.
Fetch an RSS, Atom or JSON Feed address live and validate it. Returns whether it is valid, the feed title and website link, format, number of items, date of the newest item, how many days ago that was, whether the feed is still active (an item in the last 90 days), the publishing pace, and the errors and warnings found (HTTP status, redirects, content type, XML problems, missing dates or GUIDs). Use when someone asks "does this feed work", "is this RSS valid" or "is this blog still updated". Give it the feed address; for a website address use discover_feeds first.
Feed address, e.g. "https://www.theverge.com/rss/index.xml".
Pick your client. Values in capitals, like YOUR_TOKEN, are yours to fill in.
Run in a terminal
claude mcp add --transport http mcp https://checkrss.com/mcp
Fetch an RSS, Atom or JSON Feed address live and validate it. Returns whether it is valid, the feed title and website link, format, number of items, date of the newest item, how many days ago that was, whether the feed is still active (an item in the last 90 days), the publishing pace, and the errors and warnings found (HTTP status, redirects, content type, XML problems, missing dates or GUIDs). Use when someone asks "does this feed work", "is this RSS valid" or "is this blog still updated". Give it the feed address; for a website address use discover_feeds first.
| Argument | Type | Description |
|---|---|---|
| url* | string | Feed address, e.g. "https://www.theverge.com/rss/index.xml". |
{
"type": "object",
"$schema": "http://json-schema.org/draft-07/schema#",
"required": [
"url"
],
"properties": {
"url": {
"type": "string",
"maxLength": 2000,
"minLength": 3,
"description": "Feed address, e.g. \"https://www.theverge.com/rss/index.xml\"."
}
},
"additionalProperties": false
}Find the RSS, Atom or JSON feeds of a website. Reads the page for <link rel="alternate"> feed tags, knows the feed patterns of big platforms (YouTube channels and playlists, Reddit, Medium, Substack, GitHub, Tumblr, Bluesky, Stack Overflow, Apple Podcasts) and tries common feed paths such as /feed and /rss.xml when the page lists none. Returns the candidate feed addresses with format and title. Use when someone asks "does this site have RSS".
| Argument | Type | Description |
|---|---|---|
| website_url* | string | Website or page address, e.g. "https://example.com" or "youtube.com/@channel". |
{
"type": "object",
"$schema": "http://json-schema.org/draft-07/schema#",
"required": [
"website_url"
],
"properties": {
"website_url": {
"type": "string",
"maxLength": 2000,
"minLength": 3,
"description": "Website or page address, e.g. \"https://example.com\" or \"youtube.com/@channel\"."
}
},
"additionalProperties": false
}Read the most recent items of an RSS, Atom or JSON feed, fetched live. Returns up to 20 items with title, link, publication date and a short plain text summary (HTML removed). Use to answer "what did this site publish lately" or to preview a feed before subscribing.
| Argument | Type | Description |
|---|---|---|
| url* | string | Feed address, e.g. "https://hnrss.org/frontpage". |
| limit | integer | How many items, 1 to 20 (default 10). |
{
"type": "object",
"$schema": "http://json-schema.org/draft-07/schema#",
"required": [
"url"
],
"properties": {
"url": {
"type": "string",
"maxLength": 2000,
"minLength": 3,
"description": "Feed address, e.g. \"https://hnrss.org/frontpage\"."
},
"limit": {
"type": "integer",
"default": 10,
"maximum": 20,
"minimum": 1,
"description": "How many items, 1 to 20 (default 10)."
}
},
"additionalProperties": false
}List the categories and subcategories of the checkrss.com feed directory with the number of sites in each and the checkrss.com page of the category, plus the number of sites per country. Use before search_directory to pick a category or subcategory id.
No arguments.
{
"type": "object",
"$schema": "http://json-schema.org/draft-07/schema#",
"properties": {}
}Search the checkrss.com directory of 1,600+ sites whose feeds the CheckRSS bot fetched and verified, including about 300 Polish sites (national and regional news, tech, business, culture). Match by name, domain, subject or country; filter by category and country. Each result has the site, its feed addresses with format, item count and newest item date, and a checkrss.com page link. Use to recommend feeds: "RSS feeds about AI", "Polish news RSS", "BBC RSS feed". Leave query empty to list the best known sites of a category or country.
| Argument | Type | Description |
|---|---|---|
| limit | integer | How many results, 1 to 25 (default 10). |
| query | string | Words to match, e.g. "security", "bbc", "krakow". May be empty. |
| country | string | Country name or ISO code, e.g. "Poland", "PL", "UK", "Germany". |
| category | string | Category id from list_categories, e.g. "technology", "news". |
| subcategory | string | Subcategory id from list_categories, e.g. "ai", "poland-regional". |
{
"type": "object",
"$schema": "http://json-schema.org/draft-07/schema#",
"properties": {
"limit": {
"type": "integer",
"default": 10,
"maximum": 25,
"minimum": 1,
"description": "How many results, 1 to 25 (default 10)."
},
"query": {
"type": "string",
"default": "",
"maxLength": 120,
"description": "Words to match, e.g. \"security\", \"bbc\", \"krakow\". May be empty."
},
"country": {
"type": "string",
"maxLength": 60,
"description": "Country name or ISO code, e.g. \"Poland\", \"PL\", \"UK\", \"Germany\"."
},
"category": {
"type": "string",
"maxLength": 40,
"description": "Category id from list_categories, e.g. \"technology\", \"news\"."
},
"subcategory": {
"type": "string",
"maxLength": 40,
"description": "Subcategory id from list_categories, e.g. \"ai\", \"poland-regional\"."
}
},
"additionalProperties": false
}We read every tool's name, description and inputs the way the model will, looking for instructions hidden from you. A clean result means no warning signs, not a guarantee. Scanned Nov 18, 2026.