Skip to main content
The rss MCP adapter fetches and parses RSS 2.0 and Atom feeds from public URLs. No OAuth or API keys are required for typical public feeds — pass feed URLs directly in workflow steps for news monitoring, content pipelines, and aggregation.

​
Prerequisites

  • Publicly reachable RSS or Atom feed URLs (HTTPS recommended)
  • project_contributor access in AgentRuntime
  • No API credentials required for standard public feeds

​
Connect in AgentRuntime

1

Identify feed URLs

Collect RSS or Atom URLs from blogs, YouTube channels, podcasts, or product changelogs. Most sites expose /feed, /rss, or <link rel="alternate"> in HTML.
2

Create a connection (optional)

On Connections, click New custom connection only if you want to customize user_agent or request_timeout_seconds. Many teams install the adapter with default settings and no secrets.
3

Add rss MCP instance

Go to MCP, click Add instance for rss, wire a connection on Instance config if you created one, then set the profile active.
4

Test a feed

Run mcp_call → rss_fetch_items with a known feed URL and limit: 5.
See Authentication for binding connections to MCP instances.
Feeds behind login walls or IP allowlists may fail unless you expose a public URL or proxy. This adapter does not store feed credentials.

​
What you can build

  • Morning digest — rss_fetch_items_multi on industry feeds → LLM summary → email or Slack (when available).
  • Content pipeline — New blog item → LLM rewrite → human_task → publish via LinkedIn or Google Docs connectors.
  • Competitive monitoring — Daily fetch → diff titles against Postgres history → alert on new product posts.
  • Podcast/show notes — Parse episode feeds → extract links → add to ClickUp content calendar.

​
Tools

ToolDescription
rss_fetch_itemsFetch one feed URL; return normalized items sorted newest first
rss_fetch_items_multiFetch multiple feeds in parallel; merge and sort newest first

​
Example

Fetch the five latest items from a blog feed: 
{
  "id": "rss-fetch",
  "type": "mcp_call",
  "name": "Latest blog posts",
  "tool_name": "rss_fetch_items",
  "tool_args": {
    "rss_feed_url": "https://example.com/blog/feed.xml",
    "limit": 5
  },
  "timeout_s": 30
}

​
Configuration

KeyRequiredDescription
user_agentNoHTTP User-Agent for feed requests
request_timeout_secondsNoPer-feed HTTP timeout in seconds

​
Troubleshooting

IssueFix
Parse errorsConfirm URL returns XML, not HTML error page; validate feed in a reader
Empty itemsFeed may be valid but stale; set limit to 0 to return all parsed entries
TimeoutIncrease request_timeout_seconds or workflow timeout_s for slow hosts
SSL or redirect failuresUse canonical HTTPS feed URL; some sites redirect HTTP → HTTPS