Files that outlive the chat
Your AI writes into a real cloud workspace. Close the conversation, open a new one tomorrow, and the project is still there.
Cloud workspaces for AI development
Give your AI a real place to build.
VibeNAS gives AI assistants persistent project storage over MCP — and, when you need it, a managed environment that can safely run your build.
- Free to start
- Files persist
- Scoped MCP access
- Managed execution
developer-workspace
// connected over MCP
assistant@vibenas
$ update the homepage and verify the build
Read project files
Updated components
Ran TypeScript validation
Saved to persistent storage
Managed execution complete
Developer workloads run through the isolated VibeNAS execution broker.
MCPNode.jsPythonTypeScriptESLintCloud storage
What you get
Move your AI past the chat window.
Stop pasting files back and forth. Give your assistant a workspace with exactly the capabilities you choose to turn on.
A whole project, not a snippet
Source, config, docs, generated assets, nested folders. Structured the way you'd structure it on disk.
Scoped credentials
Every workspace gets its own protected credential. One account cannot read another account's workspace.
Code that actually runs
On Developer, your AI can run Node.js, Python, TypeScript, lint, tests, and builds through managed execution.
Sandboxed by default
Execution is isolated, resource-bounded, and queued — held apart from platform services and every other workspace.
Room to grow into
Start with persistent files. Add managed execution. Move to dedicated infrastructure when you need the whole machine.
Three steps
Signed up to shipping, in about a minute.
Create your workspace
Sign up. VibeNAS provisions persistent storage for your account automatically — nothing to configure.
Connect your AI
Paste your VibeNAS MCP endpoint into any compatible AI client and authenticate with your workspace credential.
Start building
Your AI creates and edits files right away. Upgrade to Developer when you want it running your toolchain too.
Your MCP endpoint
https://mcp.vibenas.com/mcp
Protected by your workspace credential
Security
Nothing here is trusted by default.
Developer workspaces use managed execution, not an open shell. Every command goes through a broker built to limit what a shared workload can reach.
Read the security model- Workspace-scoped authentication
- Cross-workspace access controls
- Per-workspace Linux execution identity
- Outbound network blocked on shared compute
- Bounded memory, CPU, task, and runtime limits
- Central execution audit events
Pricing
Start with files. Add compute. Go dedicated.
AI Workspace
Free
A persistent cloud workspace for AI-assisted projects.
Free
- 1 persistent workspace
- 500 MB storage
- Secure MCP endpoint
- Create, read, edit, move, and search files
- Workspace dashboard
- File browser and storage visibility
Managed Compute
Developer
For AI workflows that need to work with your code, not just write it.
$5/month
- Everything in Free
- 5 GB storage
- Managed Node.js execution
- Managed Python execution
- TypeScript and lint tooling
- Tests and managed builds
- Hardened shared execution sandbox
- Per-workspace queue protection
Full Development
Dedicated
Your own infrastructure, with direct system access and none of the shared-environment limits.
Early access
- Dedicated compute instance
- SSH access
- Full Git workflows
- Outbound Internet access
- Package installation
- Ports and background services
- Advanced MFA controls
- Higher resource limits
Shared Developer workspaces deliberately exclude SSH, Git networking, package installation, exposed ports, and background services.
Dedicated workspace
When the sandbox stops being enough.
Your own compute environment, with a real terminal and the traditional development workflow you already know.
Request dedicated access- SSH
- Git
- Packages
- Networking
- More resources
- Advanced MFA
Give your AI a workspace, not another attachment.
Create a free workspace, connect over MCP, and keep every project intact from the first file to the final build.