About
Unrestricted filesystem and persistent shell MCP server
README
Isolated, observable, autonomous bots that excel at memory and self-improvement. Tools, machines and coding environments are connected via MCP.
Installation
Install & Connect to Klex Cloud:
You can self-host a Klex Bot and connect it to the Klex Cloud.
Start here: https://cloud.klex.bot and locally install the Klex Bot when prompted.
Why connect it to cloud?
The cloud provides OOTB app-integrations with bot-owned identities and push-notification support (E.g. @My-Bot-123 in Slack which wakes up on a DM or @mention). Your local Klex Bot still owns all the data and API keys for inference providers.
Without the cloud, you need to set up reverse-proxies like tailscale or ngrok to receive webhooks from integrations that should wake up your bot. Most of the integrations only exist in cloud, though!
Or install just the local Klex Bot:
macOS and Linux:
curl -fsSL https://raw.githubusercontent.com/stagewise-io/klex/main/install.sh | sh
Windows (PowerShell):
irm https://raw.githubusercontent.com/stagewise-io/klex/main/install.ps1 | iex
After installation, you can add models, providers and mcp servers via the Terminal UI or configure their entries in {data-directory}/config.json directly.
Sending messages to the bot requires a connected MCP server with support for mcp-extension-push-notifications.
Principles
A Klex Bot has one permanent session
All incoming messages enter the same durable default session. Wherever you chat with the bot, it has the same context. Extensions can explicitly create isolated child sessions when a background-task needs separate context, tools, or model behavior.
Klex bots are mcp-native
The core agent loop is built around mcp-support. Incoming messages are sent via mcp servers that implement mcp-extension-push-notifications. Realtime calling is implemented via mcp-extension-realtime-media.
Klex bots can't touch their host computer
During runtime, the Klex Bot process only accesses data in one dedicated, isolated data-directory on its host machine. Connections to outside environments and machines are established via MCP. By implementing MCP extensions, MCP servers can wake-up a Klex Bot (mcp-extension-push-notifications) or initiate a voice- and video-call (mcp-extension-realtime-media).
Klex bots learn and self-improve (and keep their learnings organized)
After every unit of work, a Klex Bot will reflect on its actions and conversations - it will create and organize learnings as skills to do the task more efficiently the next time and remember preferences.
You can voice-call Klex Bots
Voice conversations via latest realtime models are integrated out of the box. mcp-extension-realtime-media lets you initiate a call via MCP - and swill make the bot switch to an agreed-upon realtime protocol (so far, only LiveKit is supported).
Community
License
Klex is licensed under the Apache License 2.0.
Install Machine in Claude Desktop, Claude Code & Cursor
unyly install machineInstalls into Claude Desktop, Claude Code, Cursor & VS Code — handles npx, uvx and build-from-source repos for you.
First time? Get the CLI: curl -fsSL https://unyly.org/install | sh
Or configure manually
Run in your terminal:
claude mcp add machine --env KLEX_MACHINE_OMIT_HOST_DETAILS="" -- npx -y @klex/machineStep-by-step: how to install Machine
FAQ
Is Machine MCP free?
Yes, Machine MCP is free — one-click install via Unyly at no cost.
Does Machine need an API key?
Yes, it requires environment variables: KLEX_MACHINE_OMIT_HOST_DETAILS. Unyly injects them into the config during install.
Is Machine hosted or self-hosted?
Self-hosted: the server runs locally on your machine via the install command above.
How do I install Machine in Claude Desktop, Claude Code or Cursor?
Open Machine on unyly.org, pick your client tab (Claude Desktop, Claude Code, Cursor) and press Install — the config is generated automatically, no JSON editing.
Related MCPs
GitHub
PRs, issues, code search, CI status
by GitHubFilesystem
Secure file operations with configurable access controls.
Memory
Knowledge graph-based persistent memory system.
Unyly Deploy
Deploy sites and MCP servers to Unyly straight from your AI agent
by UnylyAmap Maps Mcp Server
MCP server for using the AMap Maps API
by duxiaohuiPayMCP
Lightweight payments layer for MCP servers: turn tools into paid endpoints with a two-line decorator. [PyPI](https://pypi.org/project/paymcp/) · [npm](https://w
by blustAIServer For Power BI
Enables exploring Microsoft Fabric / Power BI workspaces and semantic models, and executing ad-hoc DAX queries.
by Mitsubishi-FusoGpt Image 2 Mcp
RunAPI GPT Image 2 MCP server for image generation: create tasks, poll results, and check pricing across 1 model from Claude Code, Codex, Cursor, and VS Code.
by runapi-aiTemplate MCP Server
A CLI tool to create a new Model Context Protocol server project with TypeScript support, dual transport options, and an extensible structure
by mcpdotdirectSupabase
Database, auth and storage
by SupabaseCompare Machine with
Not sure what to pick?
Find your stack in 60 seconds
Author?
Embed badge for your README
Browse similar
All development MCPs
