Click on "Deploy Server".
Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state.
In the chat, type @ followed by the MCP server name and your instructions, e.g., "@Local-Files-View-MCP list the files in the project folder"
That's it! The server will respond to your query, and you can continue using it as needed.
Here is a step-by-step guide with screenshots.
An MCP server that allows ChatGPT and other AI assistants to access and inspect local files.
Related MCP server: filesystem-mcp
Start the MCP server:
uv run server.py --root /path/to/sandboxAdd the following endpoint to Claude Desktop's MCP configuration:
http://127.0.0.1:8000/mcpSet sandbox root to the directory you want Claude to be able to access.
Start the MCP server using Streamable HTTP:
uv run server.py --root /path/to/sandboxOnce the server is running, connect your MCP client to the following endpoint:
http://127.0.0.1:8000/mcpIf you want to use the server with services such as browser-based ChatGPT that cannot directly connect to localhost, you will need to expose the HTTP endpoint using a service such as ngrok.
When exposing the server publicly, make sure that everyone who can connect to the endpoint is trusted to access the files under the sandbox root.
This MCP server is intended for personal use.
If the server is exposed over HTTP, users who can connect to the endpoint may be able to read files under the sandbox root.
Do not expose the endpoint to untrusted users.
Also, do not set the sandbox root to a directory containing sensitive or confidential information.
This server cannot be deployed
Securely search and manage workspace context files for AI agents and teams.
Give your AI agent a persistent map of your project's structure, dependencies, and bugs.
Sandbox workspace tools: search, file read, DB queries, integrations. Returns synthetic data.