nono
Secure, kernel-enforced sandbox CLI and SDKs for AI agents. Capability-based isolation with secure key management, atomic rollback, cryptographic immutable audit chain of provenance. Run your agents in a zero-trust environment.
claude mcp add --transport stdio always-further-nono npx -y @modelcontextprotocol/server-filesystem /data
How to use
This MCP server entry integrates with the Model Context Protocol tooling to host an MCP server that exposes filesystem-backed resources for agent interactions. By leveraging the nono sandboxing and policy framework, it enables controlled access to a specified data directory while enforcing kernel-level sandbox guarantees. Clients (agents) can connect to this server to browse, read, and interact with the supported resources according to the server’s policies, while guards and rollbacks ensure that any destructive actions are blocked or rolled back as configured.
To use the server, start it with the MCP command and target a data path that should be exposed to clients. The CLI examples show running an MCP server via npx, which will install and execute the server package on demand. Once running, you can direct agents or other MCP clients to interact with the server by addressing the server’s endpoint and passing the appropriate commands or requests defined by the MCP spec for the filesystem server. The server is designed to be composable with other MCP components, enabling complex permission models and audit trails across multiple servers.
How to install
Prerequisites:
- Node.js (with npm) installed on the host machine
- Internet access to fetch MCP packages via npx
Installation steps:
-
Install Node.js if it is not already installed. You can download it from https://nodejs.org/ and verify with: node -v npm -v
-
Run the MCP server using npx (the server will be installed on-the-fly if needed): npx -y @modelcontextprotocol/server-filesystem /data
-
Verify the server starts and is reachable by the MCP client tooling or by querying its endpoint as per your environment.
Optional:
- If you have a predefined server image or package, you can pin the version in the npx command, e.g. npx -y @modelcontextprotocol/server-filesystem@1.2.3 /data
Additional notes
Tips and notes:
- The server name used here is aligned with the filesystem MCP server example package. If your deployment uses a different package, replace the npm package name accordingly in the command.
- The -y flag in the npx command auto-confirms installation prompts; omit it if you prefer interactive prompts.
- Ensure that the target path (e.g., /data) exists and has the appropriate permissions for the MCP server to access.
- Review security policies and environment configuration to enforce the desired sandbox rules and audit trails.
- If you run into network or compatibility issues, verify that your Node.js environment supports npx and that your network allows fetching npm packages at install time.
Related MCP Servers
bytebot
Bytebot is a self-hosted AI desktop agent that automates computer tasks through natural language commands, operating within a containerized Linux desktop environment.
cursor-talk-to-figma
TalkToFigma: MCP integration between AI Agent (Cursor, Claude Code) and Figma, allowing Agentic AI to communicate with Figma for reading designs and modifying them programmatically.
ios-simulator-skill
An IOS Simulator Skill for ClaudeCode. Use it to optimise Claude's ability to build, run and interact with your apps, without using up any of the available token/context budget.
memov
Give git-like & traceable memory to OpenClaw and any coding agents. By https://memov.ai/ aka Entire CLI for every coding agents by MCP.
eion
Shared Memory Storage for Multi-Agent Systems
MCP-OpenStack-Ops
Professional OpenStack operations automation via MCP server. Specialized tools for cluster monitoring, instance management, volume control & network analysis. FastMCP + OpenStack SDK + Bearer auth. Claude Desktop ready. Perfect for DevOps & cloud automation.