cbt-query 0.1.0

CBT Query MCP Server with cross-platform support

CBT Query MCP Server

A simple Model Context Protocol (MCP) server for querying TensorRT test coverage and case mapping data.

Features

• Query all test cases and files
• Get coverage mapping by case name
• Query cases by files and/or functions
• Simple HTTP client with proper error handling
• Minimal logging and clean code structure

Installation

Prerequisites

• Python 3.10 or later
• pip package manager

Option 1: Using Installation Scripts

Windows

# Run the Windows installation script
install.bat

Unix/Linux/macOS

# Run the Unix installation script
chmod +x install.sh
./install.sh

Option 2: Manual Installation

Windows

# Create and activate virtual environment
python -m venv venv
venv\Scripts\activate.bat

# Install the package
pip install -e .

Unix/Linux/macOS

# Create and activate virtual environment
python3 -m venv venv
source venv/bin/activate

# Install the package
pip install -e .

Option 3: Using uv (if available)

# Initialize with uv
uv init
cd cbt-query

# Create virtual environment and activate it
uv venv
source .venv/bin/activate  # Unix
# or
.venv\Scripts\activate.bat  # Windows

# Install dependencies
uv add "mcp[cli]" httpx requests

Usage

Running the Server

After installation, you can run the server in several ways:

Method 1: Using the installed command

# Activate virtual environment first
source venv/bin/activate  # Unix
# or
venv\Scripts\activate.bat  # Windows

# Run the server
cbt-query

Method 2: Using Python module

python -m cbt_query

Method 3: Direct execution

python main.py

Configuration for Cursor

Add the following configuration to your ~/.cursor/mcp.json file:

Windows

export CBT_SERVER_URL="http://dlswqa-nas:12345/"

{
    "mcpServers": {
        "cbt_query": {
            "command": "python",
            "args": [
                "-m", "cbt_query"
            ],
            "cwd": "C:\\path\\to\\your\\cbt-query\\installation"
        }
    }
}

Unix/Linux/macOS

{
    "mcpServers": {
        "cbt_query": {
            "command": "python3",
            "args": [
                "-m", "cbt_query"
            ],
            "cwd": "/path/to/your/cbt-query/installation"
        }
    }
}

Alternative configuration using uv

{
    "mcpServers": {
        "cbt_query": {
            "command": "uv",
            "args": [
                "--directory",
                "/path/to/your/cbt-query",
                "run",
                "python",
                "-m", "cbt_query"
            ]
        }
    }
}

Available Tools

The MCP server provides the following tools:

• query_all_cases: Get all test cases from the server
• query_all_files: Get all files from the server
• query_by_case: Get coverage mapping by case name
• query: Query cases by files and/or functions

API Examples

# Get all cases
await query_all_cases()

# Get all files
await query_all_files()

# Get coverage by case
await query_by_case("test_case_name")

# Query by file
await query(file_name="example.cpp")

# Query by function
await query(funcs="example_function")

# Query by file and function
await query(file_name="example.cpp", funcs="example_function")

Development

uv init new_project cd new_project

Create virtual environment and activate it

uv venv source .venv/bin/activate

Install dependencies

uv add "mcp[cli]" httpx requests pip

cursor setup

# ~/.cursor/mcp.json
{
  "mcpServers": {
    "trt_query": {
      "command": "python",
      "args": [
        "-m",
        "cbt_query"
      ],
      "env": {
        "CBT_SERVER_URL": "http://dlswqa-nas:12345/"
      }
    }
  }
}

Environment Setup

Make sure to set the CBT_SERVER_URL environment variable:

export CBT_SERVER_URL="http://your-server:12345"

Troubleshooting

Common Issues

1. CBT_SERVER_URL not set: Make sure the environment variable is set
2. Import errors: Verify all dependencies are installed in the active environment
3. Connection errors: Check that the CBT server is running and accessible

Logging

The server logs to stderr with INFO level by default.