Q4_learning is an MCP server designed to support Quarter 4 Agentic AI learning workflows. It leverages the MCP (Model Context Protocol) to manage model context, prompts, and tool utilization in a structured way, enabling researchers to run experiments, manage prompt templates, and orchestrate agent tasks. The server exposes an interface for loading and evaluating agent policies, executing defined tasks, and exchanging context-rich messages with model backends. Use it to run prompt engineering experiments, track context changes across iterations, and coordinate tool use (e.g., search, reasoning, or execution tools) within a consistent MCP framework.

To use the server, start it with the configured runtime (Python in this case). Once running, you can connect clients that implement MCP-compatible endpoints to send prompts, request tool invocations, and receive context-aware responses. Typical workflows include loading a prompt template, attaching a task-specific context (such as current quarter goals or assignment constraints), and iterating on agent-based solutions. Ensure you review any environment variables (like host/port and log level) to tailor the server to your environment.

Prerequisites:

• Python 3.8+ installed on your system
• Access to install Python packages (pip)
• Git installed to clone the repository (optional if you already have the code locally)

Installation steps:

1. Clone the repository (or prepare your workspace): git clone <repository-url> cd Q4_learning

2. Create and activate a virtual environment (recommended): python -m venv venv

   Windows

   venv\Scripts\activate.bat

   macOS / Linux

   source venv/bin/activate

3. Install dependencies (if a requirements.txt or pyproject.toml exists):

   If a requirements.txt is present

   pip install -r requirements.txt

   If using Poetry (optional alternative)

   poetry install

4. Run the MCP server (as defined in mcp_config):

   If using the default module approach

   python -m mcp_server

5. Verify the server is running by checking logs or visiting the configured host/port (e.g., http://localhost:8000).

Notes:

• If no requirements.txt is present, install core dependencies as needed for your MCP implementation (e.g., FastAPI, asyncio utilities, or any MCP-specific libraries).
• You may need to adjust environment variables for host, port, or logging before starting the server.

Tips and common issues:

• Ensure the port you configure (MCP_PORT) is open and not used by another service.
• If your MCP server uses a specific module name (e.g., mcp_server), verify the module path is correct and that it's importable from your running environment.
• For debugging, set LOG_LEVEL to DEBUG to gain verbose insight into prompt handling and tool invocation.
• If you plan to run multiple MCP servers, consider using distinct host/port pairs and separate virtual environments.
• When transitioning prompts or contexts between iterations, ensure context serialization is compatible with the MCP protocol to avoid data loss.