Feedly
Utilities & PersonalGive agents a monitored view of the web through Feedly. They subscribe to feeds, pull stream contents and entries, tag and save articles, and mark items read, turning your Feedly boards into a research pipeline.
What This Integration Enables
Feedly is where a team's view of the outside world accumulates, and the failure mode is always the same: the feeds are good, the folders are organized, and nobody has time to read them. This connector turns those streams into a research pipeline that runs whether or not anyone opens the app. FlowRunner agents pull entries from feeds, folders, and global streams, hydrate full content efficiently by fetching ids first and batching the retrieval, then summarize, score, file, and route what they find. - Pull articles from any stream: a single feed, a folder, or everything at once - Page large streams efficiently with entry ids first, bulk hydration second - Summarize and classify articles with AI agents, then route the results - Save and tag articles into boards that other tools and teammates read from - Manage the subscription list itself: subscribe, unsubscribe, keep unread state honest The connector cannot spend money or delete data beyond the account's own reading state, so the human's role is editorial rather than protective, and FlowRunner's [human-in-the-loop](/concepts/human-in-the-loop) step lands on the judgment calls: what deserves the team's attention, and what the curated record should contain.
Without FlowRunner
With FlowRunner
Use Case Scenarios
The daily digest that reads 40 feeds so nobody has to
Each morning the agent walks the industry folder with Get Stream Entry IDs, hydrates via Get Multiple Entries, and produces per-article summaries with a relevance score. The top items post to [Slack](/integrations/slack) as a ranked digest with links; everything processed gets Mark As Read so the unread count reflects only what arrived since. The team's morning scan drops from an hour of skimming to five minutes of reading conclusions.
Competitive intelligence with an audit trail
Articles mentioning tracked competitors get pulled from their dedicated folder, summarized, and appended to a log in [Google Sheets](/integrations/google-sheets) with title, source, date, and link. Items an analyst confirms as significant are filed with Save Entry and Tag Entry to the competitive board, and a parallel entry lands in [Notion](/integrations/notion) where the quarterly strategy review is drafted. Six months later, the question "when did they start talking about pricing changes" has a searchable answer.
The feed list that maintains itself
Monthly, the agent audits List Subscriptions against actual output: feeds that produced nothing in 90 days are proposed for removal, and new source candidates surfaced from article citations are proposed for addition. A person approves the changes, then Subscribe To Feed and Unsubscribe From Feed execute the batch, with new feeds filed into the right categories. The monitored view of the web stays deliberate instead of archaeological.
Human-in-Loop Highlight
Tag Entry is quiet but consequential: the boards it files into are the team's curated record, the stream a leadership brief or client-facing newsletter is assembled from, and an agent that files aggressively turns a curated board into a second firehose, which defeats the entire point of curation. So the pipeline splits the work by what each party is good at. The agent reads everything, summarizes everything, and scores everything; the analyst decides what enters the board, because inclusion is an editorial claim ("this matters, and we noticed it") that carries the team's credibility with whoever reads the output downstream. The same review covers Unsubscribe From Feed during list cleanups: dropping a source is cheap to execute and expensive to discover you needed, so the pruning list gets human eyes before it runs.
Agent Capabilities
12 actionsAccount
1- Get Profile Retrieves the authenticated user's profile: user id, email, name, subscription tier, and locale. The user id is the building block for category, tag, and board stream ids.
Subscriptions
3- List Subscriptions Lists all subscribed feeds with feed id, title, website, update timestamps, and folder membership. Subscription ids become streamIds for content pulls.
- Subscribe To Feed Subscribes to an RSS feed by id, prefixing bare URLs with feed/ automatically, with an optional display title and folder assignment.
- Unsubscribe From Feed Removes a feed subscription from the account. Runs inside reviewed list cleanups, not autonomous pruning.
Streams
2- Get Stream Contents Retrieves articles from a stream: a feed, a folder, or a global stream. Supports ranking by newest or engagement, unread-only filtering, a newer-than timestamp, and cursor pagination.
- Get Stream Entry IDs Retrieves only entry ids for a stream, the efficient first step for large streams, with a continuation token for paging. Feeds Get Multiple Entries and Mark As Read.
Entries
2- Get Entry Retrieves one article's full content by id: title, author, publish date, origin feed, content, links, keywords, and engagement metrics.
- Get Multiple Entries Retrieves full content for many articles in one request. The bulk hydration step that keeps large-stream processing cheap.
Markers
2- Mark As Read Marks entries read or unread by applying a read marker, keeping unread counts meaningful after automated processing.
- Save Entry Saves or unsaves entries to the Read Later list, landing them in the global saved stream.
Tags and Boards
2- List Tags Lists the user's tags and boards with ids and labels, the collections that organize saved research.
- Tag Entry Adds an article to a tag or board, creating the tag if it does not exist. The filing step behind curated boards, and the one this page's editorial gate watches.
Frequently Asked Questions
What can FlowRunner do with Feedly?
FlowRunner agents can run Get Profile, List Subscriptions, and Subscribe To Feed in Feedly, plus 9 more actions.
Does connecting Feedly to FlowRunner require OAuth?
No. Feedly connects to FlowRunner with an API key, no OAuth flow required.
Can Feedly trigger a FlowRunner workflow automatically?
Feedly doesn't currently expose triggers in FlowRunner. It connects as an action step inside workflows started by another trigger.
Start building with Feedly
$100 in credits. No card required. Connect in minutes.