Linkhut
LinkHut manages bookmarked links in a minimalistic, shareable interface, helping teams organize URLs and track references in one place
Verdict
Common use cases
- Save links from Slack threads into shared library
- Tag and organize research URLs during planning
- Retrieve bookmarks by tag for project briefs
- Clean up outdated links in team knowledge base
- Archive meeting notes with tagged references
Integration
- Vendor
- Linkhut
- Category
- other
- Auth
- OAUTH2
- Tools
- 5
- Composio slug
linkhut
Tools
- Add bookmark
Adds a new bookmark to linkhut. the bookmark can be marked as private/public and read/unread, with optional tags and notes.
- Delete bookmarkdestructive
This tool allows users to delete a bookmark from their linkhut account by providing the bookmark url. it operates independently and only requires the url parameter to identify and remove the bookmark.
- Get all tags
Retrieves a list of all tags and their usage counts for the authenticated user. no additional parameters required besides authentication.
- Get bookmarks
This tool retrieves all bookmarks from the user's linkhut account. it makes a get request to the api endpoint and handles authentication. the tool returns a list of bookmarks including details such as url, title, tags, notes, and timestamp.
- Update Bookmark
This tool allows users to update an existing bookmark in linkhut. the tool updates the metadata of a bookmark including its title, description, and tags.
Setup
Setup guide
- 11. In Switchy, open your workspace settings and navigate to the Integrations tab. 2. Find Linkhut in the MCP directory and click Connect. 3. You'll be redirected to Linkhut's OAuth authorization page — sign in if prompted. 4. Grant Switchy permission to read and write bookmarks and tags on your behalf. 5. After authorizing, you'll return to Switchy and see Linkhut listed as connected. 6. Open any Space and type '@Linkhut get bookmarks' to confirm the connection works. 7. If you see your bookmark list, you're ready — the MCP can now add, update, delete, and search links in response to @mentions.
What teammates see: by default, memories from Linkhut are scoped to the Space (PROJECT visibility) - you can mark any memory PRIVATE or share it ORG-wide.
Works well with
Top models
Compatibility data appears once enough Spaces have used this MCP together with a given model.
How Switchy teams use it
Starter prompts
Save This Link
@Linkhut add a bookmark for https://example.com/article with tags 'design, research' and note 'Referenced in Q2 planning'. Mark it public and unread.Open in a Space →
Find Tagged Bookmarks
@Linkhut get all bookmarks tagged 'onboarding' and list their titles and URLs.Open in a Space →
Update Bookmark Metadata
@Linkhut update the bookmark at https://example.com/guide to have title 'Team Onboarding Guide' and add tag 'HR'.Open in a Space →
List All Tags
@Linkhut get all tags and show me the top 10 by usage count.Open in a Space →
Delete Outdated Link
@Linkhut delete the bookmark at https://example.com/old-post from my account.Open in a Space →
Example outputs
Illustrative - representative of the model's voice and quality, not literal recordings.
@linkhut show me all my bookmarks tagged 'design-system' that I haven't read yet
Found 7 unread bookmarks tagged 'design-system' in your Linkhut account. These include articles on component libraries (3), accessibility patterns (2), and token architecture (2). The oldest unread item is from 3 weeks ago. Most are marked public. Would you like me to summarize the top 3 by recency or help you organize them into subcategories?
This example shows Linkhut's filtering and retrieval capabilities. The MCP fetches all bookmarks, then the AI filters by tag and read status. Useful for surfacing saved resources you meant to review. Requires read scope during OAuth setup.
@linkhut save this article about React Server Components to my reading list with tags 'react' and 'architecture', mark it unread
Bookmark added to your Linkhut account. Title extracted as 'Understanding React Server Components'. Tagged with 'react' and 'architecture', marked unread and public by default. The bookmark is now accessible at linkhut.com/bookmarks and will appear in your unread queue. You can update privacy or add notes anytime.
Demonstrates the add-bookmark action. The MCP creates the entry immediately—this is a write operation that modifies your Linkhut account. If you're in a shared Space, teammates see the confirmation but can't access your private bookmarks unless you share credentials.
@linkhut look at my 20 most recent bookmarks and suggest which tags I'm overusing or should consolidate
Analyzed your recent bookmarks and tag distribution. You've used 'javascript' 8 times and 'js' 6 times in the last 20 saves—consider consolidating to one canonical tag. 'frontend' and 'ui' overlap on 4 bookmarks; merging them would simplify filtering. You also have singleton tags like 'misc-tools' and 'random' that could be retired. Want me to draft update commands to standardize these?
This synthesis example pairs Linkhut's get-bookmarks and get-all-tags tools with AI reasoning. The MCP provides raw data; the AI identifies patterns and suggests cleanup. Helpful for maintaining a tidy bookmark taxonomy over time. No account changes occur unless you approve follow-up updates.
Use-case deep-dives
When Linkhut works for async knowledge sharing
A 6-person remote design team saves links all week—articles, inspiration, competitor teardowns—then reviews them together in a Friday Slack thread. Linkhut's tagging and public/private flags let each designer curate their own feed while surfacing team-relevant finds. The OAuth2 setup is a one-time friction point, but once connected, adding bookmarks from Slack or standup notes takes seconds. This works if your team actually reviews the digest weekly. If links pile up unread for months, you're just building a graveyard. The buying call: Linkhut replaces the chaotic 'links' channel when your team commits to a review cadence.
Linkhut as a lightweight client portal alternative
A 3-person consulting shop onboards new clients with a curated set of 20-30 links: project templates, Loom walkthroughs, Figma files, Google Docs. They tag each bookmark by client name and mark them public so clients can access a simple URL. The 'update bookmark' tool keeps descriptions current as deliverables evolve. This beats a Notion page if your resources are already scattered across domains and you don't want to migrate everything into one CMS. The threshold: if you need version control or granular permissions beyond public/private, Linkhut's too simple. For straightforward link curation with minimal overhead, it closes the gap.
When bookmark tagging beats a spreadsheet
A 4-person product team monitors 15 competitors, saving feature launches, pricing changes, and blog posts as they surface. Each bookmark gets tagged with competitor name, feature category, and priority level. The 'get all tags' tool surfaces usage counts, so the PM sees which competitors are moving fastest. This works if your intel is link-heavy and you want fast retrieval by tag. If you need structured data—launch dates, pricing tiers, impact scores—a spreadsheet or Airtable is the better call. Linkhut wins when speed of capture matters more than rigid schema, and your team already thinks in tags.
Frequently asked
What does the Linkhut MCP do in Switchy?
It connects your Linkhut bookmark library to Switchy's AI workspace. Your team's AI agents can search saved links, add new bookmarks with tags and notes, mark items as read or private, and retrieve your full tag taxonomy. Useful when you want AI to reference or organise your team's curated link collection without leaving the conversation.
Does Linkhut MCP need OAuth or can I use an API key?
Linkhut MCP uses OAuth2, so you'll authorise Switchy through Linkhut's login screen rather than copying a token. The OAuth flow requests read and write access to your bookmarks and tags. You don't need admin rights on a Linkhut team account — any member can connect their own library.
Can the MCP edit bookmark URLs or move links between accounts?
No. The MCP can update a bookmark's title, description, tags, and privacy settings, but it cannot change the underlying URL. It also can't transfer bookmarks to another Linkhut account. If you need to migrate links, export from Linkhut directly and re-import them elsewhere.
Why use the MCP instead of just pasting links into chat?
The MCP lets AI search your entire bookmark archive by tag or keyword, retrieve metadata you've already written, and save new links with consistent tagging. Pasting a URL gives the AI no context about why you saved it or how it relates to other resources your team has curated.
Who on the team should connect the Linkhut MCP?
Whoever maintains your team's shared bookmark library. Each Switchy user connects their own Linkhut account, so if you use a single team Linkhut login, one person should authenticate it in Switchy and share the workspace. Individual connections work if everyone curates their own links separately.