# LinkAce MCP > LinkAce MCP Server lets your AI agent manage all your saved web links. Use it to create new bookmarks with custom descriptions, organize them into tags and collections (lists), and search your entire library using natural language commands. It connects directly to any self-hosted LinkAce instance. ## Overview - **Category:** productivity - **Price:** Free - **Tags:** bookmarking, archiving, knowledge-base, link-management, tagging, search ## Description **The `create_new_bookmark` tool** lets your agent drop a new web link into your LinkAce archive just by giving it the URL. It doesn't stop there; you can tell your AI client to assign a title and even write a custom description for that link right away, keeping track of what the page was about when you saved it. **The `create_new_collection` tool** lets your agent build brand-new organizational folders—the kind that keep related links grouped up. You tell it to make one, and suddenly you've got a whole new section for all those project notes or articles on AI ethics. **The `create_new_tag` tool** handles labeling. If you wanna categorize something by topic, say 'Finance' or 'Project X,' your agent can add a fresh tag to the system so you keep everything sorted by type. When it comes to finding stuff, **the `search_bookmarks` tool** is what you need. You just ask your AI client for keywords—like "Q3 marketing strategy"—and it searches your entire link library, pulling up every bookmark that matches those terms. It's way faster than clicking through folders. Need to know what you got? **The `list_all_bookmarks` tool** gives you a complete rundown of every single link currently sitting in your account. If you wanna see all the organizational buckets you’ve set up, **the `list_all_collections` tool** pulls that list for you, and if you need to check what kind of tags are floating around, **the `list_all_tags` tool** shows every tag available. For deep dives, **the `get_bookmark_details` tool** lets your agent fetch the specific metadata—like the original title or a custom description—for any single bookmark ID you point it toward. It’s perfect for checking on one link without having to browse through everything else in the archive. Finally, when a link is dead or useless, **the `delete_bookmark` tool** lets your agent scrub an existing URL out of the whole LinkAce archive permanently. ## Tools ### create_new_bookmark Adds a new link URL to your LinkAce archive, requiring only the web address. ### create_new_collection Creates a brand-new list or folder for grouping bookmarks. ### create_new_tag Adds a new category label (tag) to your bookmark system. ### delete_bookmark Removes an existing link from your entire LinkAce archive. ### get_bookmark_details Fetches specific metadata, like the description or original title, for a single bookmark ID. ### list_all_bookmarks Retrieves a list of every link currently stored in your LinkAce account. ### list_all_collections Lists all the organizational collections (lists) you have created. ### search_bookmarks Searches and returns links from your archive that match specified keywords. ### list_all_tags Shows a full list of every tag currently used to label your bookmarks. ## Prompt Examples **Prompt:** ``` Add 'https://www.wikipedia.org' to my LinkAce bookmarks. ``` **Response:** ``` I've added Wikipedia to your LinkAce archive. The new bookmark ID is 'link-123' and it is currently marked as private. ``` **Prompt:** ``` Search my LinkAce library for 'Artificial Intelligence'. ``` **Response:** ``` I found 5 bookmarks matching 'Artificial Intelligence', including links to research papers and industry news. Would you like to see the full details for any of them? ``` **Prompt:** ``` List all my bookmark collections. ``` **Response:** ``` I've retrieved your collections. You have 3 active lists: 'Reading List', 'Technical Docs', and 'Work Projects'. ``` ## Capabilities ### Bookmark Creation Your agent adds a new URL to your LinkAce archive and can assign it titles and descriptions. ### Collection Management The agent creates whole groups of bookmarks (collections) or lists, keeping related links separated. ### Tagging Links You can assign multiple tags to a bookmark, allowing you to categorize it by topic or type. ### Library Search The agent searches your entire link archive using keywords and natural language queries. ### Link Retrieval You can list all bookmarks, check details for a specific link ID, or get a complete list of available tags and collections. ## Use Cases ### The After-Hours Research Dump You spend a day reading 30 articles, saving them all manually. At 10 pm, you're overwhelmed by the sheer volume. You ask your agent: 'Take these five URLs and put them in a new collection called 'Project Alpha'. Tag them with 'review' so I remember to look at them.' The agent runs `create_new_collection` and assigns tags using multiple tools, solving the organization bottleneck. ### The Targeted Deep Dive You are writing a report on AI ethics. You don't know where you saved that one perfect link from last month. Instead of running `list_all_bookmarks` and scrolling through hundreds, you ask the agent: 'Search my bookmarks for links containing both 'AI' and 'ethics'.' The agent runs `search_bookmarks`, delivering only relevant results. ### The Documentation Audit Your dev team keeps dumping random documentation snippets. You need to clean up the mess. You ask your agent: 'List all my current collections and tags.' The agent runs `list_all_collections` and `list_all_tags`, giving you a quick overview of what needs pruning. ### The Link Cleanup Crew Your bookmarks list is bloated with dead links. You ask your agent to check the details for 10 specific bookmark IDs using `get_bookmark_details`. This lets you verify if they still point to active content before deciding which ones to remove via `delete_bookmark`. ## Benefits - Keep everything organized without manual effort. Instead of manually creating a tag or list, your agent handles it when you ask the AI client to 'File these links under 'DevOps' and mark them as 'must-read'.' (Uses `create_new_tag` and `create_new_collection`). - Stop losing context. When you find an article that’s useful but not for your current project, use the agent to immediately run `create_new_bookmark`. It handles the archiving step instantly. - Get immediate answers from years of saved research. Instead of opening 15 different folders and manually searching, just ask the agent to 'Search my bookmarks for anything related to quantum computing' (Uses `search_bookmarks`). - Full control over your data. You can run `list_all_collections` or `list_all_tags` to see exactly how your knowledge base is structured before you start adding stuff. - Maintain a clean system. If a link is outdated, don't keep it cluttering up your library. Use the agent to trigger `delete_bookmark` and archive maintenance. ## How It Works The bottom line is you give your agent the keys and the address; everything else happens through conversation. 1. Subscribe to this server on the Vinkius Marketplace. 2. Enter your personal LinkAce API Key (Bearer Token) and your Instance URL credentials. 3. Start prompting your AI client. You can then tell it, 'Search for all links tagged 'research' that were added last month.' or 'Add this article to my 'DevOps' collection.' ## Frequently Asked Questions **How do I use the search_bookmarks tool with LinkAce MCP Server?** You prompt your agent naturally. Just say, 'Search my links for anything related to Kubernetes.' The agent runs `search_bookmarks` and returns results matching the keywords. **Does create_new_bookmark require a URL?** Yes, it requires at least a valid URL. You can't add a bookmark without knowing what page it points to. **What is the difference between list_all_collections and list_all_bookmarks?** `list_all_collections` shows you the container folders (the lists). `list_all_bookmarks` shows you every single individual link inside all those containers. **Can I create a new tag using the agent?** Yes. If you tell your agent to categorize something, it runs `create_new_tag` first if that tag doesn't exist yet. **How do I authenticate when running the `create_new_bookmark` tool?** You must provide your LinkAce API Key as a Bearer Token. Your agent uses this token for every request, ensuring secure access to your self-hosted instance URL. **What happens if I try to use `delete_bookmark` on an ID that doesn't exist?** The server returns a specific 404 error code. Your agent should catch this standard HTTP response and inform the user that the bookmark ID isn't found, preventing the process from failing completely. **What metadata does `get_bookmark_details` return for a link?** It returns the full data package for that single link. You get the URL, title, description, creation date, and all associated tags or collections attached to it. **If I use `list_all_bookmarks` and have thousands of entries, how does the server handle the request?** The API uses standard pagination. You won't get everything in one call; you must pass a page size limit or cursor to fetch the data in manageable chunks. **How do I find my LinkAce API Token?** Log in to your LinkAce dashboard, navigate to User Settings, and generate a new API token from the API section. **What is the Instance URL?** This is the full base URL where your LinkAce instance is hosted, for example: `https://links.yourdomain.com`. **Can I search for specific tags?** While you can search for links by keyword, you can also list all tags to find the specific label IDs used for organization.