MCP Café
Simulates technical discussions with various agent profiles to help solve engineering problems through coffee breaks, walks, or showers.
README
MCP Café
Sometimes, when one is stuck on an engineering problem, it helps to take a coffee break and talk to friends and co-workers about it. This is what MCP Café is all about.
MCP Café is an MCP server that simulates technical discussions. Different agents can be configured to simulate different personalities and profiles.
Profiles
<!-- profiles start -->
- Senior Principal Engineer: A senior engineer with a lot of experience, who can provide insights and advice on complex problems.
- Enthusiast and Creative Intern: A young and eager intern who is enthusiastic about learning and exploring new ideas. He can think outside the box and come up with creative solutions.
- Senior Researcher: A senior researcher with a deep understanding of the latest technologies and trends. He can provide insights into the latest research and development in the field.
- Bean Counter Project Manager: A project manager who is focused on the accounting and financial aspects of the project. He can remind you of the budget and deadlines, and help you stay on track.
- Senior Software Architect: A senior software architect who can provide insights into the design and architecture of the system. He can help you understand the big picture and how different components fit together.
- Workaholic Junior Programmer: A child-free junior programmer who spends all his time coding. He is very focused on the technical aspects of a project and can help you with coding-related questions.
- PhD Student: A PhD student who is the expert on a tiny part of a related field. He can provide creative and innovative solutions to problems, or suggest new directions for research.
- Somewhat Burned Out Sysadmin: A quite negative and pessimistic sysadmin who is always worried about the worst-case scenario. He can help you think about potential problems and how to avoid them.
- Talented but Uncontrollable Rock star Engineer: A talented and creative engineer who is not afraid to challenge the status quo. He can provide innovative solutions and push the boundaries of what is possible.
- The White Hat Hacker: A security expert who can help you think about security and privacy issues. He can provide insights into the latest security trends and best practices.
- The UX Designer: A user experience designer who can help you think about the user interface and user experience of the system. He can provide insights into the latest design trends and best practices.
- The Boss: The boss is a person down to earth, who wants on time, on spec, and on budget delivery. He is fair, but expects results. He can help you stay focused on the project goals and ensure that you are meeting the requirements.
<!-- profiles end -->
You are free to create other profiles that match more closely your needs.
Available Tools
| Tool | Description |
|---|---|
have_a_coffee |
Have a coffee at the office to relax and think about the problem. Simulates conversations with various office profiles. |
go_for_a_walk |
Go for a walk to clear your mind. Provides internal reflection and introspection to help think through problems. |
take_a_shower |
Take a shower to refresh your mind and be in the best bug-fixing environment. Similar to walking but focused on finding solutions. |
Installation
-
Clone the repository:
git clone https://github.com/your-username/mcp-cafe.git cd mcp-cafe -
Create a virtual environment and install dependencies:
python -m venv .venv source .venv/bin/activate # On Windows: .venv\Scripts\activate pip install -r requirements.txt -
Make sure you have Ollama installed and running with the
gemma3model (or setMCP_CAFE_MODELto your preferred model). -
Run the server:
./start.sh
Usage example in mcp.json
{
"servers": {
// ...
"mcp-cafe": {
"type": "stdio",
"command": "/absolute-path-to-mcp-cafe/start.sh",
"args": [],
"env": {
"MCP_CAFE_OLLAMA_ENDPOINT": "http://localhost:11434",
"MCP_CAFE_MODEL": "gemma3",
// "MCP_CAFE_OLLAMA_TOKEN": "token-if-needed"
}
},
// ...
},
}
Configuration
MCP Café is designed to be generic, and can work with any Chat LLM. It can work with ChatGPT, Claude, Gemini, Mistral, Gwen, and many others. However, currently only Ollama is supported out of the box.
The default model is gemma3, but you can change it to any model you want by setting the MCP_CAFE_MODEL environment variable.
You can configure the Ollama endpoint by setting the MCP_CAFE_OLLAMA_ENDPOINT environment variable. The default is http://localhost:11434.
Moreover, if your Ollama server requires an API token, you can set the MCP_CAFE_OLLAMA_TOKEN environment variable. It will be then used for bearer token authentication.
Licence
Café is licensed under the WTFPL licence.
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.