--- title: "Pinboard Integration" description: "Pinboard is the no-nonsense bookmarking service. Agents add, update, and delete bookmarks, manage tags, keep notes, and read recent or dated bookmark sets for research and archiving flows." url: https://flowrunner.ai/integrations/pinboard date_modified: 2026-09-04T14:52:04-07:00 --- # Pinboard [Productivity](https://flowrunner.ai/integrations/category/project-management-productivity) Pinboard is the no-nonsense bookmarking service. Agents add, update, and delete bookmarks, manage tags, keep notes, and read recent or dated bookmark sets for research and archiving flows. [Verified](https://flowrunner.ai/integrations/verified "What does verified mean?") · 15 actions · API key · available [Pinboard website](https://pinboard.in/) · [Platform Documentation](https://pinboard.in/api) · Capability data verified 2026-08-24 1. A link worth keeping arrives, from a monitoring alert, a newsletter, or a teammate in chat 2. Get Bookmarks is called with the URL to find out whether the archive already holds it 3. The agent pulls the page title and summary, and Suggest Tags returns both the tags the account already uses and the popular ones others applied to the same URL 4. The agent maps the suggestions onto the taxonomy the account actually maintains rather than inventing a new label 5. Add Bookmark saves it with the resolved tags, with replacement switched off so an existing entry is reported instead of silently overwritten 6. A weekly digest built from Get Recent Bookmarks and Get Bookmark Dates goes to the team 7. Before any tag is renamed or removed across the archive, a person approves the merge plan, because folding two tags together cannot be undone ## What This Integration Enables Pinboard is a deliberately small service, and the API says so: every method is a GET, writes included, the credential travels as a query parameter, and success and failure both come back as HTTP 200 with the outcome buried in the body. FlowRunner's connector reads that result code and raises on failure, which matters more here than usual, because a workflow that only checks status treats a refused write as a completed one. The rate limits are strict and set per method rather than per account, so a well behaved workflow checks Get Last Update before it downloads anything and stops when the timestamp has not moved. What Pinboard offers in exchange for that austerity is an archive with no lifecycle management and no expiring links. FlowRunner agents save bookmarks with resolved tags, read the archive by day, by recency, or in full, keep the tag taxonomy tidy, and read the account's notes. The design commitment worth understanding is that a Pinboard tag is global. It is not a folder and not a per item label; renaming it or deleting it reaches every bookmark that carries it at once. That single fact decides where the [human-in-the-loop](https://flowrunner.ai/concepts/human-in-the-loop) gate belongs on this connector. Every action is built and verified against Pinboard's published v1 API. ### Without FlowRunner **Tags drift into synonyms**: The same idea accumulates four labels because whoever saved the link typed what came to mind **Saving is manual and skipped**: Capturing a source is a small task with no owner, so most of what a team reads is never kept **Overwrites happen silently**: Re-saving a URL replaces the existing entry by default, taking its tags and description with it ### With FlowRunner **One taxonomy, applied consistently**: Suggestions are checked against the tags the account already uses before anything is written **Capture is a step in the workflow**: The link is saved where it was found, with a title, a description, and tags already on it **Existing entries are recognized**: The agent looks the URL up first and reports a collision rather than quietly replacing what was there ## Use Case Scenarios ### Capture at the point the link appears A competitor publishes, a monitoring alert fires, or somebody drops a URL into a [Slack](https://flowrunner.ai/integrations/slack) channel the team uses for research. The agent picks it up, fetches the title and a short summary, and calls Suggest Tags, which returns two lists: tags drawn from everyone who bookmarked that URL, and tags drawn from the account's own history. It prefers the second, because that is the taxonomy the team actually reads by. Add Bookmark saves it with replacement switched off, so a URL already in the archive comes back as a collision rather than quietly overwriting the note somebody wrote on it last year. The link is kept in the moment it mattered, not in the weekly tidy up that never happens. ### A reading digest built from the archive, not from memory Once a week the agent calls Get Last Update first. If the timestamp has not moved, it stops, which is exactly what Pinboard asks callers to do. When it has, the agent uses Get Recent Bookmarks for the week's additions and Get Bookmark Dates to see which days were dense, then groups the results by tag and writes a digest into [Notion](https://flowrunner.ai/integrations/notion) or a sheet in [Google Sheets](https://flowrunner.ai/integrations/google-sheets), with a short summary posted to the team channel. Nobody has to remember what they saved. The archive reports on itself. ### A research brief assembled from a tag Somebody is writing a positioning piece and needs everything the team has kept on a topic. The agent pulls the account's tag list with Get Tags to confirm the exact label, reads the matching bookmarks, and pulls any long form context out of the account's notes with List Notes and Get Note. Because each note carries a hash of its own text, the agent can tell whether a note changed since it last read it without comparing bodies. The output is a brief with sources attached, dropped into a [Google Docs](https://flowrunner.ai/integrations/google-docs) document for the writer to work from. ## Human-in-Loop Highlight The costly operation on Pinboard is not deleting a bookmark. It is Rename Tag. Renaming a tag to a name that already exists does not fail; it folds the two tags together across the entire archive, which is how duplicates get merged, and there is no operation that unmerges them. Delete Tag has the same reach, stripping a tag from every bookmark that carries it in one call. A tidy up agent can find the drift easily enough, since Get Tags returns every tag with the number of bookmarks using it. What it should not do is act on what it found. So it prepares the plan and stops: "Four tags look like the same idea: `positioning` (61), `positioning-strategy` (9), `messaging` (34), `msg` (3). Folding the small three into the large two would merge 12 bookmarks and cannot be reversed. Approve this merge, change the target names, or leave it alone?" A person answers, the agent runs exactly the renames that were approved, and it writes the resulting mapping into the digest so the team knows what their index looks like now. The sweep found the mess. The person decided what counts as the same idea. Agent processes routinely Detects exception requiring judgment Clear match Continues automatically Ambiguous Routes to human via preferred channel Human decides Agent resumes with decision ## Agent Capabilities 15 actions ### Bookmarks 7 - **Get Last Update** Returns when a bookmark was last added, changed, or deleted. Pinboard asks that this be checked before any bulk read, because an unchanged timestamp means there is nothing to download. - **Add Bookmark** Saves a bookmark. Replacement is on by default, so saving a URL already in the archive overwrites it silently; turning it off makes Pinboard report the collision instead. - **Delete Bookmark** Deletes a bookmark by its URL. A URL that was never bookmarked comes back as not found rather than quietly succeeding, and the connector raises that. - **Get Bookmarks** Returns the bookmarks from one day, or the single one matching a URL. This is the lookup a save flow uses to check for an existing entry. - **Get Recent Bookmarks** Returns the most recently saved bookmarks. Capped at a hundred, and Pinboard permits this call only once a minute. - **Get All Bookmarks** Returns every bookmark in the account. Pinboard permits this once every five minutes and asks that Get Last Update be checked first, so a mirror workflow should be built to skip it rather than to schedule it. - **Get Bookmark Dates** Returns every day on which bookmarks were saved with a count for each, which is a cheap way to find the days worth reading given that Get Bookmarks works one day at a time. ### Tags 4 - **Suggest Tags** Returns tag suggestions for a URL: popular ones drawn from everyone who bookmarked it, and recommended ones drawn from the account's own tags. The second list is the one a consistent taxonomy comes from. - **Get Tags** Returns every tag in the account with how many bookmarks use it. The read that makes taxonomy drift visible. - **Delete Tag** Removes a tag from every bookmark that carries it. The bookmarks survive; the label is gone from all of them at once. - **Rename Tag** Renames a tag across every bookmark. Renaming into an existing tag folds the two together, which is the merge this page's human gate exists for. The old name is matched without regard to case, and an empty new name does nothing. ### Notes 2 - **List Notes** Returns the account's notes with their titles, lengths, and a hash of each body. The text itself is not in the listing. - **Get Note** Returns one note with its full text. Its hash means a workflow can tell whether the note changed without comparing bodies. ### Account 2 - **Get RSS Key** Returns the account's secret RSS key, which is what makes private Pinboard feeds readable. Treat the result as a credential. - **Get API Token** Returns the token half of the account's API credential. Worth knowing that requesting a new token in Pinboard's settings invalidates the previous one, which is the usual reason a working connection stops working. ## Frequently Asked Questions ### What can FlowRunner do with Pinboard? FlowRunner agents can run Get Last Update, Add Bookmark, and Delete Bookmark in Pinboard, plus 12 more actions. ### Does connecting Pinboard to FlowRunner require OAuth? No. Pinboard connects to FlowRunner with an API key, no OAuth flow required. ### Can Pinboard trigger a FlowRunner workflow automatically? Pinboard doesn't currently expose triggers in FlowRunner. It connects as an action step inside workflows started by another trigger. **Work at Pinboard?** This integration exposes Pinboard to AI agents on every FlowRunner plan, including through MCP, at no cost to you. [See what FlowRunner offers integration partners](https://flowrunner.ai/integrations/partners), including how to keep this page current. --- Markdown version of https://flowrunner.ai/integrations/pinboard. Site index: https://flowrunner.ai/llms.txt