MCP-dev
Custom MCP tools for local development that enable file system access, shell command execution, and web resource browsing. Designed for local use with safety considerations for potentially destructive operations.
README
MCP-dev
Some custom MCP tools for local use. By Martin Dittus in Sept 2025.
https://github.com/dekstop/MCP-dev
Install
Checkout to a local dir.
Set up a virtualenv, activate it, then install the dependencies:
python3 -m venv .venv --prompt MCP-dev
source .venv/bin/activate
python3 -m pip install -r requirements.txt
Setup
Register the MCP server in your MCP client.
To be safe, it is recommended to only auto-allow non-destructive tool calls -- tool calls like create_file, run_shell_command, and open_link are best treated with caution and reviewed before every call. But you do you.
LM Studio
Click Show Settings, select the Program tab, click the Install dropdown, select Edit mcp.json. Add the following:
{
"mcpServers": {
"MyCustomMCPTools": {
"command": "sh",
"args": [
"/path/to/MCP-dev/run.sh"
]
}
}
}
Click Save. In the Program tab a new MCP server has now appeared under the name mcp/my-custom-mcp-tools, click its toggle to enable it globally. Alternatively, enable it selectively per chat via the Integrations menu in the chat message form.
Optional: add system prompt
Add something like the following to your query context, e.g. your Preset system prompt in LM Studio:
Access web resources for any questions that require up to date information. Prioritise official sources such as Wikipedia, BBC, the Financial Times, and others. Prioritise facts over speculation. Don't consult sources that are known to exaggerate, drive scandal, or cause division.
You can access the local file system. Assume a MacOS directory structure.
You can create python scripts and run them. Any scripts and other files need to be saved in the sandbox directory. Don't install any new python libraries. You can run commands to test if a python library is installed.
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.
E2B
Using MCP to run code via e2b.
Neon Database
MCP server for interacting with Neon Management API and databases
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.