How is Conduit different from other remote connection managers?
Conduit ships a built-in MCP server that exposes every connection and credential operation as tools for AI agents — no other manager offers this.
@advenimus
Cross-platform remote connection manager (SSH, RDP, VNC, web) with a built-in MCP server exposing 60+ tools. Connect Claude Code or Codex to your infrastructure and let AI agents execute commands in SSH terminals, take screenshots of RDP/VNC desktops, drive web sessions, and read
使用下面的配置,将此服务器添加到你的 MCP 客户端。
{
"mcpServers": {
"conduit": {
"command": "node",
"args": [
"/Applications/Conduit.app/Contents/Resources/mcp/dist/index.js"
],
"env": {
"CONDUIT_SOCKET_PATH": "/Users/YOUR_USERNAME/Library/Application Support/conduit/conduit.sock",
"CONDUIT_ENV": "production"
}
}
}
}工具列表暂未提供
当前收录暂时没有可展示的工具列表,不代表服务没有工具。请查阅项目文档了解支持的能力。
工具是从 README 中自动提取的。维护者可以在 ## Tools 标题下列出工具,即可填充这部分内容。
Conduit Desktop is a cross-platform remote connection manager for SSH, RDP, VNC, and web sessions, packaged with an encrypted credential vault and a built-in Model Context Protocol (MCP) server. It allows AI agents like Claude Code or Codex to drive all connection, session, and credential operations through tool calls, making it an infrastructure bridge for DevOps, SREs, sysadmins, and AI-first engineers.
Install the app from GitHub Releases or conduitdesktop.com, then register its MCP server with your AI agent. For Claude Code, run a command like claude mcp add conduit -- node "/Applications/Conduit.app/Contents/Resources/mcp/dist/index.js" --env CONDUIT_SOCKET_PATH="..." --env CONDUIT_ENV="production" (paths vary by OS). After setup, ask your agent to list connections, run commands, or interact with remote desktops — the agent communicates with Conduit over a local Unix socket or Windows named pipe.
Conduit ships a built-in MCP server that exposes every connection and credential operation as tools for AI agents — no other manager offers this.
No. Conduit does not include its own AI. You bring your own Claude Code or Codex subscription, and Conduit never proxies your API calls.
All credentials live in a local AES-256 encrypted vault. Cloud sync is optional and uses zero-knowledge encryption; keys never leave the client.
The MCP server communicates over a local Unix socket (or named pipe on Windows). Conduit never sees your API keys, tokens, or conversation content.
Yes. The entire desktop client — including the Electron app, protocol handlers, FreeRDP helper, MCP server, and credential vault — is Apache 2.0 licensed.

Your own LinkedIn, WhatsApp, Instagram, Telegram, Email and Calendar accounts, usable from any agent — search, read and act on your own accounts.

All-in-one agile project management built for AI agents and people. Over MCP your agent works across every module: board, backlog, roadmap, whiteboards, retros, metrics and wiki.
Meeting transcripts for AI agents: search every call, read who said what, pull decisions and action items, get quotes from one person across meetings, follow deals and projects, and transcribe files or YouTube links. Hos

TripLedger is a personal travel ledger for AI assistants. Connect Claude, ChatGPT, Cursor or any MCP client to log journeys, stays and travel expenses (including from booking emails), then ask for tax-residency day count

One photo of a building problem becomes a Triage Profile in about ten seconds: what the asset is, how urgent it is (EMERGENCY, TODAY, THIS WEEK or WHENEVER), which trade to call, and what to tell the resident. For proper
评论