Capture from anywhere on the web. Cryptie auto-tags from the page context and the vocabulary you already use—so a library of thousands stays organized without filing.
Then plug it into the rest of your stack. Claude, ChatGPT, Obsidian, MCP, n8n, and any HTTP client share the same library through one REST API and Bearer token. Search it with tag: and domain: in the app, or let your agent ask for it directly.
AI auto-tagging • REST API + MCP • Claude / ChatGPT / Obsidian ready • Chrome extension • Built for huge libraries

Clean interface with powerful search, tags, and link previews
Tiago Forte's second brain runs on four steps: capture, organize, retrieve, and express. Link Crypt does all four for the URLs you collect — and now your AI assistants share the same library.
Chrome extension, iOS share sheet, web app, or POST /links from an agent — one click puts a URL into the same crypt.
Built-in AI reads page context and proposes tags from your existing vocabulary, so new saves slot into your system instead of piling up untagged.
Combine tag:ai, domain:github.com, and title: filters. Built for libraries in the hundreds or thousands.
REST API + MCP server give Claude, ChatGPT, Obsidian, n8n, and anything HTTP read/write access — one Bearer token, same library.
One-click saving from any webpage. Add tags, notes, and organize as you browse—works seamlessly across all your devices.
Add to Chrome — It's Free
People who collect lots of links and want them back on demand — surfaced by their AI when it matters, not buried in folders.
Source lists, interview refs, background reading — tag by beat or story, then run domain: or tag: when you're on deadline.
Competitors, specs, user research — keep everything labeled and pull tag:competitor or domain: in one search.
Datasets, papers, dashboards — large libraries stay usable when filters combine (e.g. tag:ai domain:arxiv.org).
Obsidian for notes, Claude / ChatGPT for reasoning, Link Crypt for URLs — wire the API and your agents save, search, and tag links without copy-paste.
Sign up with email or magic link, install the Chrome extension, and pull in your old bookmarks from Chrome, Safari, Firefox, Edge, Pocket, Instapaper, or 10+ other services.
AI auto-tagging reuses your existing tag vocabulary on every new save. Browse and refine with tag: and domain: filters as the library grows.
One Bearer token wires your library into every tool you already think with.
The second brain idea (Tiago Forte) is an organized, searchable store of everything you capture from the web. In the AI era, that store is also the context your agents read. Link Crypt is the URL layer of a modern second brain: capture from the Chrome extension or any HTTP client, organize with Cryptie's auto-tagging, retrieve with tag: / domain: search, and express by giving Claude, ChatGPT, Obsidian, and any agent live read/write access to the same library.
Link Crypt is an AI-organized link library built for retrieval at scale: Cryptie auto-tags every save from your existing vocabulary, you query with tag:ai domain:github.com, and a full REST API + MCP server lets Claude, ChatGPT, Obsidian, and your agents read and write the same library. Classic read-later apps optimize for in-app reading; we optimize for finding and feeding your AI.
Built-in bookmarks break down when you have hundreds or thousands of URLs: shallow search, awkward folders, no portable query syntax, and no API for agents or automation. Link Crypt gives you structured tags, tag:/domain: search, cross-device sync, and programmatic access — so large libraries stay usable.
Anyone who saves a lot of links and needs them back fast — journalists and writers, product and UX folks, analysts and researchers, and people wiring links into Obsidian, agents, or a second brain. If "where did I save that?" is a recurring question, Link Crypt is built for you.
Yes! Link Crypt supports imports from Pocket, Instapaper, Raindrop.io, Pinboard, GoodLinks, Readwise, browser bookmarks, and 10+ other services. All tags and metadata are preserved during import. Exporting from your browser? Follow the HTML bookmark export guide.
Use precise filters like tag:javascript, domain:github.com, title:tutorial, or description:api. Combine filters for surgical precision: tag:ai domain:arxiv.org finds all AI research papers instantly.
Absolutely. No social features, no ads, no data mining. Your links stay completely private. Export to CSV or JSON anytime—we'll never hold your data hostage.
Yes — the full REST API is on every plan, plus an MCP server (beta) for AI clients. Step-by-step setup guides: Claude & Claude Code, ChatGPT & Custom GPT, Obsidian + AI, plus the full integrations directory (n8n, Zapier, Make, Raycast, Notion, Readwise). One Bearer token from Preferences, one library, everything stays in sync.
tag: and domain: filters that scale to thousands of links, and express by giving Claude, ChatGPT, Obsidian, MCP clients, and any agent live read/write access through the REST API.tag:keyword (e.g. tag:ai), domain:website (e.g. domain:github.com), title:text, description:text, and combined queries like tag:ai domain:github.com. Imports from Pocket, Instapaper, and other services preserve tags. A full REST API and MCP server fit second-brain stacks (Obsidian) and AI agent workflows (Claude, Claude Code, ChatGPT, Custom GPT, n8n, Zapier, Manus, and more).tag:/domain: query syntax, and no API for assistants or automation. Link Crypt adds structured tags, fast filtered search, sync across devices, export, and programmatic access — so big libraries stay manageable.