Local AI Files MCP
Provides safe, allowlisted local filesystem access for MCP-compatible AI clients, enabling directory listing, file reading/writing (with byte limits), and text search while blocking paths outside approved roots and symlink escapes.
README
Local AI Files MCP
Allowlisted local filesystem access for MCP-compatible AI clients.
This project gives desktop AI clients a controlled way to inspect and edit files on your device. It exposes a local MCP server over stdio for desktop clients such as Claude Desktop, and a Streamable HTTP mode for remote/tunneled clients such as ChatGPT custom MCP apps.
What It Can Do
- Show configured allowed folders.
- Inspect file and directory metadata.
- List directories.
- Read UTF-8 text files with byte limits.
- Search text files with optional glob filtering.
- Create, overwrite, or append text files.
The server refuses every path outside explicit allowed roots, resolves real paths to block symlink escapes, and enforces read/write byte limits.
Install
npm install
npm run build
Allow at least one folder:
node dist/cli.js allow ~/Projects
Run With Claude Desktop
Build the project, then add this to Claude Desktop's MCP config, replacing the path:
{
"mcpServers": {
"local-ai-files": {
"command": "node",
"args": [
"/home/hemanth/local-ai-files-mcp/dist/cli.js",
"stdio"
]
}
}
}
Restart Claude Desktop and ask it to use local-ai-files.
Build Claude Desktop Extension
The included manifest.json follows MCPB manifest version 0.3.
npm run pack:mcpb
That creates local-ai-files-mcp.mcpb. Install it in Claude Desktop through Settings -> Extensions -> Advanced settings -> Install Extension.
Run HTTP Mode For ChatGPT-Style Remote MCP
ChatGPT web cannot connect directly to a local MCP server. Run HTTP mode locally, then expose it through a secure tunnel or OpenAI Secure MCP Tunnel if available for your workspace.
node dist/cli.js http --host 127.0.0.1 --port 3033
MCP endpoint:
http://127.0.0.1:3033/mcp
For ChatGPT custom apps, configure the externally reachable tunnel URL as the MCP server endpoint. Keep the tunnel private, authenticated, and scoped to your own account or workspace.
Commands
local-ai-files-mcp allow <folder>
local-ai-files-mcp deny <folder>
local-ai-files-mcp roots
local-ai-files-mcp stdio [folder...]
local-ai-files-mcp http [folder...] --host 127.0.0.1 --port 3033
Passing folders to stdio or http allows them for that process without changing the saved config.
Security Notes
- Do not allow your whole home folder unless you really want the AI to see it.
- Prefer project-specific folders.
- Treat
write_fileas a real local write operation. - Do not expose HTTP mode to the public internet without authentication and network controls.
- Search skips common heavy folders such as
.git,node_modules,dist, andbuild.
Verification
npm run check
Recommended Servers
playwright-mcp
A Model Context Protocol server that enables LLMs to interact with web pages through structured accessibility snapshots without requiring vision models or screenshots.
Magic Component Platform (MCP)
An AI-powered tool that generates modern UI components from natural language descriptions, integrating with popular IDEs to streamline UI development workflow.
Audiense Insights MCP Server
Enables interaction with Audiense Insights accounts via the Model Context Protocol, facilitating the extraction and analysis of marketing insights and audience data including demographics, behavior, and influencer engagement.
VeyraX MCP
Single MCP tool to connect all your favorite tools: Gmail, Calendar and 40 more.
graphlit-mcp-server
The Model Context Protocol (MCP) Server enables integration between MCP clients and the Graphlit service. Ingest anything from Slack to Gmail to podcast feeds, in addition to web crawling, into a Graphlit project - and then retrieve relevant contents from the MCP client.
Kagi MCP Server
An MCP server that integrates Kagi search capabilities with Claude AI, enabling Claude to perform real-time web searches when answering questions that require up-to-date information.
Neon Database
MCP server for interacting with Neon Management API and databases
E2B
Using MCP to run code via e2b.
Exa Search
A Model Context Protocol (MCP) server lets AI assistants like Claude use the Exa AI Search API for web searches. This setup allows AI models to get real-time web information in a safe and controlled way.
Qdrant Server
This repository is an example of how to create a MCP server for Qdrant, a vector search engine.