gomcptest is a proof of concept (POC) demonstrating how to implement a Model Context Protocol (MCP) with a custom-built host to play with agentic systems. It provides tools for building and testing AI agents that can interact with their environment through function calling.
MCP Protocol Integration
gomcptest implements the Model Context Protocol (MCP) for building custom agentic systems that can interact with tools and their environment.
Check the architecture documentation for more details!
Contributions welcome!
We do a Pull Request contributions workflow on GitHub. New users are always welcome!
Extensible Tools
Create custom tools with our extensible architecture and API compatibility layers to build your own agents.
Check the tools reference for more information.
Documentation Structure
Our documentation follows the Divio Documentation Framework, organizing content into tutorials, how-to guides, reference, and explanation.
Tutorials
Learning-oriented content that takes you through a series of steps to complete a project. Perfect for beginners getting started with gomcptest.
Start with our Getting Started tutorial.
How-To Guides
Problem-oriented content that guides you through the steps to address specific use cases and tasks with gomcptest.
Learn how to create a custom tool.
Reference
Technical descriptions of the gomcptest components, APIs, and tools with comprehensive details.
Check our Tools Reference.
Explanations
Understanding-oriented content that explains concepts and provides context about how gomcptest works.
Read about the MCP Protocol.
OpenAI-compatible server
A server that implements the OpenAI API interface and connects to Google’s Vertex AI for model inference.
Located in host/openaiserver.
cliGCP
A command-line interface similar to Claude Code or ChatGPT that interacts with Gemini models and MCP tools.
Located in host/cliGCP.
MCP Tools
Various tools that enable agent functionality:
- Bash, Edit, GlobTool, GrepTool
- LS, Replace, View
- dispatch_agent
Located in the tools directory.