Feedly MCP Server
Feeds, Boards, Read Later and read state in your Feedly, worked by your agent.
Hosted MCP Server
npx add-mcp 'https://cotlet.dev/mcp/v1/feedly'Installs into Claude Code, Codex, Cursor and more
Documentation
Cotlet · Integrations · How it works · Security · For developers · For automation agencies · Blog · API docs · Help center · Changelog · Contact · Pricing · Sign up
Feedly API and MCP server: Let your AI agent read your Feedly for you.
Feedly's own help center says "Official API access is Enterprise only" and "There is no official way to connect an AI assistant to a personal Feedly account" (updated 23 Oct 2026). Cotlet runs 14 Feedly tools hosted, over MCP for agents and REST for scripts, in the Feedly account you read in every day.
Feeds, Boards, Read Later and read state in your Feedly, worked by your agent. Use it in ChatGPT, Claude or any AI agent over MCP, or in n8n, Make and scripts over REST.
- Free 7-day trial, no card
- Unofficial, not affiliated with Feedly
- Page: https://cotlet.dev/integrations/feedly-api-mcp-server
Cotlet is an independent service. It is not affiliated with, endorsed by or sponsored by Feedly. Feedly is a trademark of its owner. If Feedly's own integrations cover what you need, use them first.
What can you automate in Feedly?
- Summarize a Feed's unread articles
- Save articles that match a rule to a Board
- Mark a week of backlog as read
- Rank what you saved for Read Later
- Find and follow sources on a topic
- Move sources between Feeds
Does Feedly have an MCP server or an API?
- The official MCP server, Feedly Threat Graph, answers questions about cyber attacks, CVEs and malware. It needs a Feedly Advanced or Enterprise plan.
- The Enterprise API reads streams, searches and adds articles to Team Boards. Its reference lists nothing for read state, Read Later or following a source.
- Zapier, Make and IFTTT need Pro+ ($144 a year). Zapier has 3 actions: Create New Article, Add Article to Board and Subscribe to Feed.
| What you need | Zapier, Make, IFTTT | Developer token and GitHub servers | Cotlet |
|---|---|---|---|
| Read a Feed's unread articles on demand | New articles as triggers | Yes | Yes, with full text |
| Mark read, keep unread | No | Yes | Yes, by ID or by age |
| Save to a Board, Read Later | Board by URL only | Not in the servers we found | Yes, Board after your OK |
| Find and follow sources on a topic | Subscribe by feed URL | Search, no follow | Yes, into a new Feed |
| Feedly plan needed | Pro+ | Any, unsupported by Feedly | Any, Free included |
| Limit | Your Zapier tasks | 50 calls a day, non-commercial | Pay per call |
Checked 27 Oct 2026. Sources: Feedly help: support knowledge, API and MCP (23 Oct 2026), Feedly help: setting up the Feedly MCP server in Claude, Feedly API: authorization, Feedly API: endpoint index, Zapier: Feedly triggers and actions, Feedly Cloud group: developer token limit
How does Cotlet connect to Feedly?
- Sign in to Feedly. Use your own account in its Cotlet Cloud Browser.
- Choose MCP or REST. Connect your agent or create a Cotlet API key.
- Run your workflow. Ask your agent or send an HTTP request.
We never keep your Feedly password, and nothing lands on a Board until you approve it.
Connect Feedly to your agent or workflow
Use MCP with your AI agent, or HTTP requests in n8n, Make, Zapier and scripts. Follow the MCP setup below or jump to the REST guide.
- Open Feedly in a Cloud Browser. Create a Cotlet account, open Feedly in its Cloud Browser, sign in with your Google, Apple or email login and press Enable for agents. Free, Pro and Pro+ all work.
- Add the Cotlet address. Paste https://cotlet.dev/mcp/v1 where your agent takes an MCP server or connector URL.
- Approve the agent. Chat apps send you to a Cotlet sign-in page. Code editors, n8n and scripts send an API key you create in the console.
Feedly shares https://cotlet.dev/mcp/v1 with every other app you enable. If you enabled Feedly after connecting, refresh the connector so the 14 tools appear.
Claude:
Settings > Connectors > Add custom connector
URL: https://cotlet.dev/mcp/v1/connect
Claude Code:
claude mcp add --transport http cotlet \
https://cotlet.dev/mcp/v1 \
--header "Authorization: Bearer $COTLET_API_KEY"
ChatGPT:
Settings > Apps & Connectors > Create
MCP server URL: https://cotlet.dev/mcp/v1/connect
Authentication: OAuth
Cursor:
// ~/.cursor/mcp.json
{
"mcpServers": {
"cotlet": {
"url": "https://cotlet.dev/mcp/v1",
"headers": { "Authorization": "Bearer ${env:COTLET_API_KEY}" }
}
}
}
Codex:
# ~/.codex/config.toml
[mcp_servers.cotlet]
url = "https://cotlet.dev/mcp/v1"
bearer_token_env_var = "COTLET_API_KEY"
n8n:
MCP Client Tool (inside an AI Agent)
Transport: HTTP Streamable
URL: https://cotlet.dev/mcp/v1
Auth: Bearer, <your key>
Tool: feedly_read
Make, Zapier:
Make: MCP Client > Call a tool
Zapier: MCP Client by Zapier > Run Tool
Server URL: https://cotlet.dev/mcp/v1
Token: <your key>
Tool: feedly_write
How to use the Feedly REST API
Create your Cotlet API key
Create a Cotlet API key with access to Feedly and its read operations. Store it as COTLET_API_KEY in your environment or automation platform's credential store.
Connect Feedly in Cotlet and enable it for agents. Account access: Your own Feedly account. Team Feeds and Team Boards appear if your account has them (Team Feeds are Enterprise only). You sign in in your own browser and keep your 2-step verification.
Make your first Feedly API request
REST:
curl https://cotlet.dev/api/v1/feedly/feeds \
-H "Authorization: Bearer $COTLET_API_KEY"
Read feeds and apply the operation limits
A successful REST call puts the tool output in result. This illustrative response excerpt omits call IDs, messages and credit details.
Illustrative response excerpt:
{
"status": "success",
"app": "feedly",
"tool": "list_feeds",
"result": {
"feeds": [
{
"feed_id": "feed_ai01",
"name": "AI news",
"type": "personal",
"unread_count": 23,
"sources": [
{
"source_id": "src_a11f02",
"title": "Model Releases Weekly",
"website": "example-models.com",
"unread_count": 14
},
{
"source_id": "src_a11f09",
"title": "Agent Tooling Digest",
"website": "example-agents.dev",
"unread_count": 9
}
]
},
{
"feed_id": "feed_ds02",
"name": "Design",
"type": "personal",
"unread_count": 10,
"sources": [
{
"source_id": "src_d3c410",
"title": "Interface Notes",
"website": "example-interface.io",
"unread_count": 6
},
{
"source_id": "src_d3c41b",
"title": "Type and Layout",
"website": "example-type.co",
"unread_count": 4
}
]
}
]
}
}
Returns every Feed in your account, personal and team, with its sources and the unread count for each. Team Feeds appear only if your account has them.
Use feed_id and source_id from this call in list_articles, follow_source and move_source.
Keep the same selection when continuing a list. Compare changed record values as well as identifiers when refreshing your destination.
Feedly workflows for agents and automations
| Tool | Connect through | Example workflow |
|---|---|---|
| n8n | HTTP Request | Articles in a Feed, source, Board or Read Later, newest first, with read state. Use list_articles in your configured workspace report. |
| Make | HTTP > Make a request | Full article text with source, link, read state and the Boards it is saved in. Use get_article in your configured workspace report. |
| Python or Node.js | REST API | Your Boards, personal and team, with the number of articles in each. Use list_boards in your configured workspace report. |
Send Authorization: Bearer <your Cotlet API key> in request headers. You configure the schedule, destination and record comparisons.
Cotlet reads Feedly; your workflow steps write to your chosen destination. These HTTP workflows do not require a native Cotlet connector or a prebuilt template.
Python and Node.js can use the same HTTP requests without an AI agent. Keep access keys in credentials, outside shared workflow exports.
The API and MCP reference lists operations and inputs. The REST API guide explains authentication, errors and approvals.
Feedly API endpoints and MCP tools
Feeds and articles
Reading a list never marks anything read. Feedly marks articles older than 30 days as read by itself, so older ones never come back as unread. Search follows your plan, and Free has none.
- List feeds (
list_feeds, read): Your Feeds with the sources in each and the unread count. - List articles (
list_articles, read): Articles in a Feed, source, Board or Read Later, newest first, with read state. - Get article (
get_article, read): Full article text with source, link, read state and the Boards it is saved in. - Search articles (
search_articles, read): Search article text in your Feeds, optionally inside one stream.
Read state and Read Later
Read state is yours alone, even in a Team Feed, so these run without asking. mark_all_read returns the IDs it changed, so one call puts them back.
- Mark articles read or unread (
mark_articles_read, write): Mark articles as read or unread, optionally hiding them from the list. - Mark a stream as read (
mark_all_read, write): Mark every article in a Feed or source as read, optionally only older ones. - Save for later (
save_for_later, write): Add an article to Read Later, or remove it.
Boards
A save to a Team Board shows the article to your team, so both tools ask first. No tool removes an article from a Board.
- List Boards (
list_boards, read): Your Boards, personal and team, with the number of articles in each. - Save article to Board (
save_to_board, write, asks first): Save an article to a Board, the star icon in the reader. - Add URL to Board (
add_url_to_board, write, asks first): Add a web page or PDF that is not in your Feeds to a Board, with a note.
Sources
Your plan's cap still applies: 100 feeds on Free, 1,000 on Pro. Unfollowing keeps what you saved from that source.
- Search sources (
search_sources, read): Search Follow Sources for websites and feeds by topic, with followers and posting rate. - Follow source (
follow_source, write): Follow a source and add it to a Feed, creating the Feed if you name a new one. - Unfollow source (
unfollow_source, write): Stop following a source. - Move source to another Feed (
move_source, write): Move a source you follow from one Feed to another.
Inputs, REST path and example response of every tool: https://cotlet.dev/integrations/feedly-api-mcp-server#tools
Who can see my Feedly session, and how do I revoke it?
- What we store. Your signed-in Feedly profile, AES-256-GCM encrypted, in Finland. No password.
- What your agent gets. The 14 Feedly tools. No cookie and no developer token.
- What the session sees. Your Feeds, Boards, Read Later, notes and highlights. Team Feeds if you have them.
- What no tool touches. Plan and billing, AI Feeds, Mute Filters, newsletters, deleting Feeds or Boards.
- If Feedly signs you out. You sign in again yourself, with Google, Apple or email.
- How to cut access. Disable for agents, reset the browser, or delete the API key. The next call fails.
Notes and highlights on your articles are readable by the session, as they are by you. Our site scores A+ on Qualys SSL Labs, securityheaders.com and Mozilla Observatory (live results).
Approvals, undo and mistakes
- Reads the new articles. One Feed since the last run, or a search across all of them.
- Checks your rule. Full text when the preview does not settle it.
- Skips what is filed. An article already on the Board is left alone.
- Waits for your OK. Each Board save shows up in your Cotlet console first.
| Action | Undo |
|---|---|
| Mark articles read | Yes, state unread on the same IDs |
| Mark a Feed or source read | Yes, the call returns the IDs to restore |
| Save for later | Yes, saved set to false |
| Follow or unfollow a source | Yes, the opposite tool |
| Move a source | Yes, move it back |
| Save to a Board, add a URL | Remove it in Feedly; no tool does |
2 of 14 tools ask first: save_to_board and add_url_to_board, because a Team Board shows the article to your team. Read state, Read Later and your sources reach no one, so those run at once. You can turn the wait off for a personal Board you file to every day.
Feedly's own Shift + A clears a whole list. The tools work on the article IDs the agent actually read, so what it skipped stays unread.
MCP call:
{
"name": "add_url_to_board",
"arguments": {
"board_id": "board_31c0",
"url": "https://example-research.org/state-of-feeds.pdf",
"note": "Chapter 3 compares polling and push."
}
}
REST:
curl -X POST https://cotlet.dev/api/v1/feedly/boards/board_31c0/url-to-boards \
-H "Authorization: Bearer $COTLET_API_KEY" \
-H "Content-Type: application/json" \
-d '{"url":"https://example-research.org/state-of-feeds.pdf","note":"Chapter 3 compares polling and push."}'
Response:
{
"board_id": "board_31c0",
"board_name": "Reading list",
"url": "https://example-research.org/state-of-feeds.pdf",
"title": "State of feeds report",
"note": "Chapter 3 compares polling and push.",
"saved": true
}
A cron job or n8n flow calls the same tools over REST at /api/v1/feedly with the same key and price, for example a morning digest of one Feed posted to Slack. A save to a board_id that does not exist fails, and failed calls are free. Each tool row above shows its REST path.
How much does the Feedly MCP server cost?
About $32 a month for one reader with about 150 sources, 5,540 credits. Your Feedly plan stays as it is; Free works.
| Job | Calls | Credits |
|---|---|---|
| Digest of one Feed, then mark it read, daily | 7 × 30 | 2,100 |
| File matching articles to a Board, each workday | 6 × 22 | 1,320 |
| Clear unread older than 7 days, weekly | 3 × 4 | 120 |
| Rank the week's Read Later, weekly | 10 × 4 | 400 |
| Find and follow sources for a new topic | 10 × 1 | 100 |
| Cloud Browser, 30 days | 1,500 | |
| Total | 5,540 |
- Free 7-day trial, no card
- 10 credits per successful call, MCP or REST
- 50 credits a day for the Cloud Browser
- Works on Free, Pro and Pro+
Feedly publishes no usage numbers, so the month is our example: about 150 sources, one daily digest, Board filing on 22 workdays, weekly cleanup and one new topic. That costs more than Feedly Pro+ itself ($144 a year); if Zapier's 3 actions cover your job, Pro+ is cheaper. Dollars at the $29 pack rate.
Feedly MCP server on GitHub, Zapier or hosted: which to use
| Option | What it does | Where it stops |
|---|---|---|
| c0ffee0wl/feedly-mcp-server | Python, GPL-3.0, 12 tools: subscriptions, unread counts, articles, mark read and keep unread. 7 stars. | Developer token; no Boards, Read Later or follow. Last push 30 Jan 2026. |
| fredrsat/feedly-mcp | Node, MIT, 6 tools: folders, feeds, articles, unread counts, source search, mark read (off by default). 0 stars. | Not released, build from source; developer token. Last push 13 Sep 2026. |
| hafnium49/mcp_server_feedly | Search, collect and entity lookup on the Enterprise API. 0 stars. | Enterprise threat intelligence only. Last push 11 Jun 2025. |
| Zapier MCP for Feedly | Create New Article, Add Article to Board and Subscribe to Feed as MCP tools. | Pro+ only; it cannot read a Feed on demand or change read state. |
| Feedly Threat Graph | Feedly's official MCP server for threat intelligence. | Advanced or Enterprise; it does not reach your own Feeds. |
Every GitHub server uses the v3 developer token. Feedly support calls that API "unsupported and undocumented" and says it "can stop at any time", and Feedly staff have put it at 50 calls a day for non-commercial use. If you would rather switch readers, Inoreader has an official API on its Pro plan. To keep the Feedly account you already run, Cotlet works there with no token at all.
Who runs Cotlet
Take Group sp. z o.o. runs Cotlet. It is a software company registered in Poland, EU (VAT PL6342984273), in business since 2019, led by its founders Bart Krawczyk (Co-founder, CTO) and Jake Slugocki (Co-founder, SEO, sales and marketing). Questions about the Feedly tools go to contact@cotlet.dev.
Feedly API and MCP: questions
How do I get access?
Sign up free at https://cotlet.dev, open Feedly under Integrations in the console and sign in to Feedly in its Cloud Browser. The trial starts when you sign up and asks for no card.
How do I connect Feedly to ChatGPT or Claude?
Sign up at cotlet.dev, log in to Feedly once in its Cloud Browser and add Cotlet to ChatGPT or Claude, following the steps in your console. No API key is needed. ChatGPT and Claude then work in your Feedly account through Cotlet, in their phone apps too.
What can ChatGPT do in Feedly?
With Cotlet, ChatGPT, Claude and other AI agents handle jobs like these in Feedly: summarize unread articles from your feeds; save relevant articles to a Board; find and follow new sources for a topic. You ask in plain words, and anything that reaches other people waits for your OK first.
Do I need Feedly's API add-on or an Enterprise plan?
No. Cotlet works in your own Feedly account and uses no API token. Feedly says self-service API tokens are only available to Enterprise clients, and that the API is an add-on you arrange with its sales team.
What happens to unread counts when my agent marks articles as read?
mark_articles_read lowers the unread count for the source and Feed, the same as Feedly's Mark as read. With hide set to true it also removes the articles from the list until you reload. In a Team Feed, read status is per person, so teammates still see their own unread articles.
Can I use a Feedly developer token instead?
You can try, with a GitHub server. Feedly support calls the v3 developer API unsupported and says it can stop at any time, and Feedly staff have limited the token to 50 calls a day for non-commercial use. Cotlet needs no token.
Can it react when a new article arrives?
Cotlet has no Feedly trigger. Run your agent on a schedule, or on Pro+ let a Zapier trigger such as New Article in Feed or New Article in Board start it.
Does it work on Feedly Free?
Yes, within what Free allows: 100 feeds and no search, so search_articles answers with an error that says so. Feedly checks Free feeds less often, every 30 minutes to a day, so a digest can lag behind the site.
Is the MCP server different from the REST API?
Only in how it is called. Both carry the same 14 tools, inputs, answers, API key and price. Agents use MCP; scripts and HTTP steps use REST.
Is Cotlet affiliated with Feedly?
No. Cotlet is independent and not affiliated with feedly, Inc.
Can I use the Feedly API without an AI agent?
Yes. Call Cotlet's REST API from a script or an HTTP step in n8n, Make or Zapier with your Cotlet key. MCP is the alternative for AI agents; you do not need to connect an agent to use REST.
Give your AI agents your signed-in browser.
Product: Integrations · How it works · Pricing · Security
Resources: Blog · API docs · Help center · Changelog · GitHub
Solutions: For developers · For automation agencies
Company: About · AI info: company facts · Contact
Legal: Terms of service · Privacy policy · Refund policy · Acceptable use · Cookie policy · DPA · Subprocessors
Elsewhere: Status page · Cotlet on GitHub · Cotlet on X · Cotlet on LinkedIn · Cotlet on Product Hunt · Cotlet on Crunchbase · Cotlet on SaaSHub
Cotlet.dev is built by Bart Krawczyk and Jake Slugocki, the founders of Take Group sp. z o.o., ul. Wincentego Janasa 4, 40-855 Katowice, Poland. KRS 0000871398, NIP 6342984273, VAT EU: PL6342984273. About us
Cotlet is an independent service. It is not affiliated with, endorsed by or sponsored by any of the products listed on this site. Product names are trademarks of their respective owners.