Pinako AI Bridge
Pinako AI Bridge is the MCP server that connects Pinako, a Chromium extension for tab, window, library, and bookmark management, to local AI clients including Claude Desktop, Claude Code, Cursor, Windsurf, Cline, Roo Code, and Continue.dev. The real payoff comes when you already use a desktop AI client: because the bridge runs inside that client alongside its other connect
README
Pinako AI Bridge
Connect your browser tabs and libraries to Claude Desktop, Cursor, and everything else your AI can reach.
The Model Context Protocol (MCP) bridge that connects Pinako — a Chromium extension for tab tree management — to local AI clients including Claude Code, Claude Desktop, Cursor, Windsurf, Cline, Roo Code, Continue.dev, and any HTTP MCP client.
The bridge runs as a small native messaging host on your computer and exposes your Pinako tab tree, libraries, library groups, Main Notes, and browser bookmarks as MCP tools. It is a full read and write surface: your AI assistant can both query your browser state and reorganize it — move and nest tabs, create libraries, groups, and folders, edit titles, tags, memos, colors, and notes, find and remove duplicates, and apply large batched edits as a single undoable step.
Because the bridge runs inside desktop AI clients like Claude Desktop, Cursor, and Codex, the same assistant can move information between Pinako and the other tools, files, and services it connects to. For example:
- "Copy this research library's notes into a Google Doc."
- "Build a spreadsheet from the tabs in my Competitors library."
- "Visit the URLs in my Reading library and write up a summary."
- "Organize my open tabs into libraries by topic, then tag the work ones."
- "Find and remove my duplicate bookmarks."
Writes go through the same tier gating, confirmation, and undo machinery as the extension's own UI. Destructive operations (deletes, bookmark reorders) return a
CONFIRMATION_REQUIREDresult until the caller explicitly confirms, and abulk_applybatch applies atomically as one undo step. Every write returns a structured{ok, error}result so clients can branch on the error code.
Tools
Read and write MCP tools. Full per-tool reference with parameters: pinako.pro/docs/ai-connect.
Read — query the tree, libraries, notes, and bookmarks, and search
get_tree, get_tree_summary, search_tabs, search_pinako, search_docs, list_libraries, get_library, get_main_tree_notes, get_bookmarks, list_browsers, find_duplicates
Tree structure — move, nest, group, and remove nodes
move_node, indent_node, outdent_node, create_group, create_folder, ghost_node, delete_node, delete_live_node, bulk_apply (up to 250 atomic sub-ops, undoable as one step)
Node metadata — titles, tags, memos, colors
set_title, set_tags, add_tags, remove_tags, set_memo, set_star_color, set_row_color
Libraries and library groups — create, edit, delete, and arrange
create_library, add_to_library, set_library_title, set_library_description, delete_library, create_library_group, set_library_group_title, set_library_group_description, delete_library_group, add_library_to_group, remove_library_from_group, reorder_library_panel, reorder_libraries_in_group
Notes — create and edit Main Notes and library notes
create_note, set_note_content, delete_note
Bookmarks — write to the browser bookmark tree
add_to_bookmarks (bookmark reads use get_bookmarks; dedup uses find_duplicates)
The read tools are safe to auto-approve. Write tools are individually annotated (idempotent, non-idempotent, or destructive) so MCP clients can gate them appropriately.
Install
Setup guide with screenshots and per-client instructions: pinako.pro/docs/ai-connect
Downloads: pinako.pro/downloads
Available installers:
- Windows — graphical installer (
.exe) - Linux x64 and Linux ARM64 — CLI installer
Requirements
- Pinako extension installed in your browser (Chrome Web Store)
- An active Pinako subscription (Pro or higher) — the bridge is gated by subscription
Build from source
git clone https://github.com/teleomorph/pinako-mcp.git
cd pinako-mcp
npm install
node setup/build.js --all
Outputs land in dist/. See setup/build.js for build flags (--all, --no-tauri, --no-cli).
The Tauri GUI installer can only be built on the platform it targets (i.e. build the Windows installer on Windows). The CLI installer cross-compiles for all listed platforms from any host.
Architecture
host.js— MCP server / Chrome native messaging hostsetup/main.js— CLI installer (writes the native messaging host JSON, configures supported AI clients)setup/build.js— multi-stage build (esbuild → pkg → Tauri)installer/src-tauri/— Rust Tauri GUI installer for Windows
License
Recommended Servers
playwright-mcp
A Model Context Protocol server that enables LLMs to interact with web pages through structured accessibility snapshots without requiring vision models or screenshots.
Magic Component Platform (MCP)
An AI-powered tool that generates modern UI components from natural language descriptions, integrating with popular IDEs to streamline UI development workflow.
Audiense Insights MCP Server
Enables interaction with Audiense Insights accounts via the Model Context Protocol, facilitating the extraction and analysis of marketing insights and audience data including demographics, behavior, and influencer engagement.
VeyraX MCP
Single MCP tool to connect all your favorite tools: Gmail, Calendar and 40 more.
graphlit-mcp-server
The Model Context Protocol (MCP) Server enables integration between MCP clients and the Graphlit service. Ingest anything from Slack to Gmail to podcast feeds, in addition to web crawling, into a Graphlit project - and then retrieve relevant contents from the MCP client.
Kagi MCP Server
An MCP server that integrates Kagi search capabilities with Claude AI, enabling Claude to perform real-time web searches when answering questions that require up-to-date information.
E2B
Using MCP to run code via e2b.
Neon Database
MCP server for interacting with Neon Management API and databases
Exa Search
A Model Context Protocol (MCP) server lets AI assistants like Claude use the Exa AI Search API for web searches. This setup allows AI models to get real-time web information in a safe and controlled way.
Qdrant Server
This repository is an example of how to create a MCP server for Qdrant, a vector search engine.