Obto MCP server
{
"mcpServers": {
"obto": {
"type": "http",
"url": "https://app.obto.co/ms/mcp"
}
}
}{
"mcpServers": {
"obto": {
"type": "http",
"url": "https://app.obto.co/ms/mcp"
}
}
}{
"mcpServers": {
"obto": {
"type": "http",
"url": "https://app.obto.co/ms/mcp"
}
}
}{
"servers": {
"obto": {
"type": "http",
"url": "https://app.obto.co/ms/mcp"
}
}
}{
"mcpServers": {
"obto": {
"type": "http",
"url": "https://app.obto.co/ms/mcp"
}
}
}file .mcp.json (in your project root)file claude_desktop_config.jsonfile ~/.cursor/mcp.jsonfile .vscode/mcp.jsonfile ~/.codeium/windsurf/mcp_config.jsonversion 3.5.62
What it does
OBTO is a full-stack deployment platform that allows AI agents to build, deploy, and run applications on a hosted workspace. Developers connect Claude, GPT, or other MCP-compatible models to describe what they want, and the AI writes the code, provisions the database, and deploys to a live URL.
- Describe a web application in plain language and have the AI build and deploy a live, functional full-stack app with frontend and backend.
- Use AI agents to handle application deployment tasks including database provisioning, code generation, and live URL assignment without manual configuration.
- Integrate OBTO with existing AI clients (Claude Desktop, Claude Web, Cursor, VS Code) to enable autonomous application development within your workflow.
Tools this server exposes
| Tool | Does |
|---|---|
| build | Initiates the AI build process for a full-stack application based on a user's description. |
| deploy | Deploys the built application to a live URL on the OBTO platform. |
| upload_large_files | Handles uploading large files during the application build and deployment process. |
| media_upload | Uploads media assets (images, videos, etc.) to the application. |
| memory | Manages persistent state and context across multiple interactions with the AI agent. |
| mcp_extension | Extends MCP capabilities for the AI agent to access additional platform features. |
| troubleshooting | Provides diagnostic tools and guidance for resolving deployment and application errors. |
Before you enable it
An MCP server runs with whatever access you give it. Start with a development account rather than production, and check which tools the server exposes before you let an agent call them unattended.