# Pythonium
A modular Model Context Protocol (MCP) server designed to enable advanced capabilities for AI agents through a robust, extensible architecture built on the official MCP SDK.
## Overview
Pythonium provides a comprehensive, production-ready foundation for building sophisticated AI agent tools and capabilities. Built around the Model Context Protocol specification and leveraging the official MCP SDK (FastMCP), it offers a clean separation of concerns through its modular package structure and streamlined configuration management.
## Architecture
### Core Packages
- **`pythonium.common`** - Shared utilities, configuration management, and base components
- **`pythonium.core`** - Central server implementation, configuration, and tool management
- **`pythonium.tools`** - Comprehensive standard tool library with extensible framework
- **`pythonium.managers`** - Lightweight management systems for specialized functionality
## Features
### MCP SDK Integration
- Built on the official MCP SDK's FastMCP framework
- Full Model Context Protocol compliance
- Multiple transport support (stdio, HTTP, WebSocket)
- Native tool registration and capability negotiation
### Comprehensive Tool Library
- **System Operations**: Command execution, environment access, system monitoring (`pythonium.tools.std.execution`)
- **File Operations**: Advanced file and directory management (`pythonium.tools.std.file_ops`)
- **Web Operations**: HTTP client, web search with multiple engines (`pythonium.tools.std.web`)
- **Tool Management**: Meta-tools for tool discovery and introspection (`pythonium.tools.std.tool_ops`)
### Advanced Configuration
- Pydantic-based configuration with validation
- Environment variable integration
- Multiple format support (YAML, JSON, TOML)
- Hot-reload capability and override support
### Production Features
- Structured logging with multiple output formats
- Comprehensive error handling and recovery
- Resource management and connection pooling
- Security considerations and rate limiting
- Performance optimizations with async architecture
## Quick Start
### Installation
```bash
# Clone the repository
git clone https://github.com/dwharve/pythonium.git
cd pythonium
# Create and activate virtual environment
python -m venv venv
source venv/bin/activate # On Windows: venv\Scripts\activate
# Install dependencies
pip install -r requirements.txt
# Install in development mode
pip install -e .
```
### Basic Usage
```bash
# Run the MCP server
python -m pythonium --help
# Start with default configuration (stdio transport)
python -m pythonium serve
# Start with HTTP transport
python -m pythonium serve --transport http --host localhost --port 8080
# Start with WebSocket transport
python -m pythonium serve --transport websocket --host localhost --port 8080
# Start with custom configuration file
python -m pythonium serve --config config/server.yaml
# Alternative: Use installed script
pythonium --help
pythonium serve
```
## Development
### Setup Development Environment
```bash
# Install development dependencies
pip install -r requirements-dev.txt
# Install pre-commit hooks
pre-commit install
# Run tests
pytest
# Run tests with coverage
pytest --cov=pythonium --cov-report=html
# Format code
black pythonium tests
isort pythonium tests
# Type checking
mypy pythonium
```
### Project Structure
```
pythonium/
├── pythonium/ # Main package
│ ├── common/ # Shared utilities and base components
│ │ ├── config.py # Pydantic configuration management
│ │ ├── base.py # Base classes and result types
│ │ ├── logging.py # Structured logging system
│ │ └── ... # HTTP client, error handling, etc.
│ ├── core/ # Core server and management
│ │ ├── server.py # Main MCP server implementation
│ │ ├── config.py # Configuration manager
│ │ └── tools/ # Tool registry and discovery
│ ├── tools/ # Tool implementations
│ │ ├── base.py # Base tool framework
│ │ └── std/ # Standard tool library
│ ├── managers/ # Specialized managers
│ ├── main.py # CLI entry point
│ └── __main__.py # Module entry point
├── tests/ # Comprehensive test suite (335 tests)
├── docs/ # Documentation
├── config/ # Configuration examples
└── requirements.txt # Dependencies
```
### Testing
The project uses pytest for testing with comprehensive coverage across all components:
- **Unit Tests**: Individual component testing
- **Integration Tests**: Cross-component interaction testing
- **Core Tests**: Configuration, server, and tool management
- **End-to-End Tests**: Full MCP server functionality
- **Performance Tests**: Load testing and benchmarks
**Current Status**: Comprehensive test coverage across all modules ensuring reliability and maintainability.
```bash
# Run all tests
pytest
# Run with coverage reporting
pytest --cov=pythonium --cov-report=html
# Run specific test categories
pytest tests/core/ # Core functionality tests
pytest tests/tools/ # Tool implementation tests
pytest tests/common/ # Common utilities tests
# Quick test run
pytest -q
# Verbose test output
pytest -v
```
## Configuration
Pythonium uses Pydantic-based configuration with support for multiple formats (YAML, JSON, TOML) and environment variable integration:
```yaml
# config/server.yaml
server:
name: "Pythonium MCP Server"
description: "A modular MCP server for AI agents"
host: "localhost"
port: 8080
transport: "stdio" # stdio, http, websocket
tools:
# Tool discovery and loading configuration
auto_discover: true
categories:
- "system"
- "web"
- "file_operations"
logging:
level: "INFO" # DEBUG, INFO, WARNING, ERROR
format: "structured" # structured, plain
output: "console" # console, file
security:
authentication: "none" # none, api_key
rate_limit:
enabled: false
requests_per_minute: 60
```
## Tool Development
### Creating a Custom Tool
```python
from pythonium.tools.base import BaseTool, ToolMetadata, ToolParameter, ParameterType
from pythonium.common.base import Result
from pythonium.common.parameters import validate_parameters
class MyCustomTool(BaseTool):
"""A custom tool example with proper parameter validation."""
@property
def metadata(self) -> ToolMetadata:
return ToolMetadata(
name="my_custom_tool",
description="A custom tool example that demonstrates the framework",
brief_description="Custom tool example",
category="custom",
tags=["example", "custom"],
parameters=[
ToolParameter(
name="message",
type=ParameterType.STRING,
description="Message to process",
required=True,
min_length=1,
max_length=1000
)
]
)
@validate_parameters # Automatic parameter validation
async def execute(self, message: str, context: ToolContext) -> Result:
"""Execute the tool with validated parameters."""
try:
result = f"Processed: {message}"
return Result.success_result(
data={"result": result, "original": message},
metadata={"tool": "my_custom_tool"}
)
except Exception as e:
return Result.error_result(f"Tool execution failed: {e}")
```
### Tool Registration
Tools are automatically discovered and registered when placed in the appropriate package structure. The tool discovery system handles:
- Automatic tool detection and registration
- Parameter validation and schema generation
- Error handling and logging
- Integration with the MCP protocol
## Contributing
We welcome contributions! Please see our [Contributing Guidelines](CONTRIBUTING.md) for details.
For a detailed history of changes, see [CHANGELOG.md](CHANGELOG.md).
### Development Workflow
1. Fork the repository
2. Create a feature branch
3. Make your changes
4. Add tests for new functionality
5. Ensure all tests pass
6. Submit a pull request
### Code Standards
- Follow PEP 8 style guidelines
- Use type hints throughout the codebase
- Maintain comprehensive test coverage
- Document all public APIs with detailed docstrings
- Use conventional commit messages
- Leverage Pydantic for data validation and configuration
- Implement proper async/await patterns for I/O operations
## License
This project is licensed under the MIT License - see the [LICENSE](LICENSE) file for details.
## Support
- **Documentation**: Available in the `docs/` directory
- **Issues**: [GitHub Issues](https://github.com/dwharve/pythonium/issues)
- **Discussions**: [GitHub Discussions](https://github.com/dwharve/pythonium/discussions)
## Acknowledgments
- Official Model Context Protocol SDK and specification
- The open-source Python community
- Pydantic and FastAPI ecosystems for configuration and validation patterns
- Contributors and maintainers
---
**Status**: Production-ready Beta - Core functionality stable, comprehensive test coverage, active development
**Current Version**: 0.1.5
**Last Updated**: July 6, 2025
**Maintainer**: David Harvey
Raw data
{
"_id": null,
"home_page": null,
"name": "pythonium",
"maintainer": null,
"docs_url": null,
"requires_python": ">=3.8",
"maintainer_email": null,
"keywords": "mcp, model-context-protocol, ai-agents, plugin-framework, server, artificial-intelligence, tools",
"author": null,
"author_email": "David Harvey <dwh.exis@gmail.com>",
"download_url": "https://files.pythonhosted.org/packages/50/7f/2ce42ddf85b5e20192e6a14ffc959ff3c1c689208b00637c021aefd93c40/pythonium-0.1.5.tar.gz",
"platform": null,
"description": "# Pythonium\n\nA modular Model Context Protocol (MCP) server designed to enable advanced capabilities for AI agents through a robust, extensible architecture built on the official MCP SDK.\n\n## Overview\n\nPythonium provides a comprehensive, production-ready foundation for building sophisticated AI agent tools and capabilities. Built around the Model Context Protocol specification and leveraging the official MCP SDK (FastMCP), it offers a clean separation of concerns through its modular package structure and streamlined configuration management.\n\n## Architecture\n\n### Core Packages\n\n- **`pythonium.common`** - Shared utilities, configuration management, and base components\n- **`pythonium.core`** - Central server implementation, configuration, and tool management\n- **`pythonium.tools`** - Comprehensive standard tool library with extensible framework\n- **`pythonium.managers`** - Lightweight management systems for specialized functionality\n\n## Features\n\n### MCP SDK Integration\n- Built on the official MCP SDK's FastMCP framework\n- Full Model Context Protocol compliance\n- Multiple transport support (stdio, HTTP, WebSocket)\n- Native tool registration and capability negotiation\n\n### Comprehensive Tool Library\n- **System Operations**: Command execution, environment access, system monitoring (`pythonium.tools.std.execution`)\n- **File Operations**: Advanced file and directory management (`pythonium.tools.std.file_ops`)\n- **Web Operations**: HTTP client, web search with multiple engines (`pythonium.tools.std.web`)\n- **Tool Management**: Meta-tools for tool discovery and introspection (`pythonium.tools.std.tool_ops`)\n\n### Advanced Configuration\n- Pydantic-based configuration with validation\n- Environment variable integration\n- Multiple format support (YAML, JSON, TOML)\n- Hot-reload capability and override support\n\n### Production Features\n- Structured logging with multiple output formats\n- Comprehensive error handling and recovery\n- Resource management and connection pooling\n- Security considerations and rate limiting\n- Performance optimizations with async architecture\n\n## Quick Start\n\n### Installation\n\n```bash\n# Clone the repository\ngit clone https://github.com/dwharve/pythonium.git\ncd pythonium\n\n# Create and activate virtual environment\npython -m venv venv\nsource venv/bin/activate # On Windows: venv\\Scripts\\activate\n\n# Install dependencies\npip install -r requirements.txt\n\n# Install in development mode\npip install -e .\n```\n\n### Basic Usage\n\n```bash\n# Run the MCP server\npython -m pythonium --help\n\n# Start with default configuration (stdio transport)\npython -m pythonium serve\n\n# Start with HTTP transport\npython -m pythonium serve --transport http --host localhost --port 8080\n\n# Start with WebSocket transport\npython -m pythonium serve --transport websocket --host localhost --port 8080\n\n# Start with custom configuration file\npython -m pythonium serve --config config/server.yaml\n\n# Alternative: Use installed script\npythonium --help\npythonium serve\n```\n\n## Development\n\n### Setup Development Environment\n\n```bash\n# Install development dependencies\npip install -r requirements-dev.txt\n\n# Install pre-commit hooks\npre-commit install\n\n# Run tests\npytest\n\n# Run tests with coverage\npytest --cov=pythonium --cov-report=html\n\n# Format code\nblack pythonium tests\nisort pythonium tests\n\n# Type checking\nmypy pythonium\n```\n\n### Project Structure\n\n```\npythonium/\n\u251c\u2500\u2500 pythonium/ # Main package\n\u2502 \u251c\u2500\u2500 common/ # Shared utilities and base components\n\u2502 \u2502 \u251c\u2500\u2500 config.py # Pydantic configuration management\n\u2502 \u2502 \u251c\u2500\u2500 base.py # Base classes and result types\n\u2502 \u2502 \u251c\u2500\u2500 logging.py # Structured logging system\n\u2502 \u2502 \u2514\u2500\u2500 ... # HTTP client, error handling, etc.\n\u2502 \u251c\u2500\u2500 core/ # Core server and management\n\u2502 \u2502 \u251c\u2500\u2500 server.py # Main MCP server implementation\n\u2502 \u2502 \u251c\u2500\u2500 config.py # Configuration manager\n\u2502 \u2502 \u2514\u2500\u2500 tools/ # Tool registry and discovery\n\u2502 \u251c\u2500\u2500 tools/ # Tool implementations\n\u2502 \u2502 \u251c\u2500\u2500 base.py # Base tool framework\n\u2502 \u2502 \u2514\u2500\u2500 std/ # Standard tool library\n\u2502 \u251c\u2500\u2500 managers/ # Specialized managers\n\u2502 \u251c\u2500\u2500 main.py # CLI entry point\n\u2502 \u2514\u2500\u2500 __main__.py # Module entry point\n\u251c\u2500\u2500 tests/ # Comprehensive test suite (335 tests)\n\u251c\u2500\u2500 docs/ # Documentation\n\u251c\u2500\u2500 config/ # Configuration examples\n\u2514\u2500\u2500 requirements.txt # Dependencies\n```\n\n### Testing\n\nThe project uses pytest for testing with comprehensive coverage across all components:\n\n- **Unit Tests**: Individual component testing \n- **Integration Tests**: Cross-component interaction testing \n- **Core Tests**: Configuration, server, and tool management\n- **End-to-End Tests**: Full MCP server functionality\n- **Performance Tests**: Load testing and benchmarks\n\n**Current Status**: Comprehensive test coverage across all modules ensuring reliability and maintainability.\n\n```bash\n# Run all tests\npytest\n\n# Run with coverage reporting\npytest --cov=pythonium --cov-report=html\n\n# Run specific test categories\npytest tests/core/ # Core functionality tests\npytest tests/tools/ # Tool implementation tests\npytest tests/common/ # Common utilities tests\n\n# Quick test run\npytest -q\n\n# Verbose test output\npytest -v\n```\n\n## Configuration\n\nPythonium uses Pydantic-based configuration with support for multiple formats (YAML, JSON, TOML) and environment variable integration:\n\n```yaml\n# config/server.yaml\nserver:\n name: \"Pythonium MCP Server\"\n description: \"A modular MCP server for AI agents\"\n host: \"localhost\"\n port: 8080\n transport: \"stdio\" # stdio, http, websocket\n\ntools:\n # Tool discovery and loading configuration\n auto_discover: true\n categories:\n - \"system\"\n - \"web\"\n - \"file_operations\"\n\nlogging:\n level: \"INFO\" # DEBUG, INFO, WARNING, ERROR\n format: \"structured\" # structured, plain\n output: \"console\" # console, file\n\nsecurity:\n authentication: \"none\" # none, api_key\n rate_limit:\n enabled: false\n requests_per_minute: 60\n```\n\n## Tool Development\n\n### Creating a Custom Tool\n\n```python\nfrom pythonium.tools.base import BaseTool, ToolMetadata, ToolParameter, ParameterType\nfrom pythonium.common.base import Result\nfrom pythonium.common.parameters import validate_parameters\n\nclass MyCustomTool(BaseTool):\n \"\"\"A custom tool example with proper parameter validation.\"\"\"\n \n @property\n def metadata(self) -> ToolMetadata:\n return ToolMetadata(\n name=\"my_custom_tool\",\n description=\"A custom tool example that demonstrates the framework\",\n brief_description=\"Custom tool example\",\n category=\"custom\",\n tags=[\"example\", \"custom\"],\n parameters=[\n ToolParameter(\n name=\"message\",\n type=ParameterType.STRING,\n description=\"Message to process\",\n required=True,\n min_length=1,\n max_length=1000\n )\n ]\n )\n \n @validate_parameters # Automatic parameter validation\n async def execute(self, message: str, context: ToolContext) -> Result:\n \"\"\"Execute the tool with validated parameters.\"\"\"\n try:\n result = f\"Processed: {message}\"\n return Result.success_result(\n data={\"result\": result, \"original\": message},\n metadata={\"tool\": \"my_custom_tool\"}\n )\n except Exception as e:\n return Result.error_result(f\"Tool execution failed: {e}\")\n```\n\n### Tool Registration\n\nTools are automatically discovered and registered when placed in the appropriate package structure. The tool discovery system handles:\n\n- Automatic tool detection and registration\n- Parameter validation and schema generation\n- Error handling and logging\n- Integration with the MCP protocol\n\n## Contributing\n\nWe welcome contributions! Please see our [Contributing Guidelines](CONTRIBUTING.md) for details.\n\nFor a detailed history of changes, see [CHANGELOG.md](CHANGELOG.md).\n\n### Development Workflow\n\n1. Fork the repository\n2. Create a feature branch\n3. Make your changes\n4. Add tests for new functionality\n5. Ensure all tests pass\n6. Submit a pull request\n\n### Code Standards\n\n- Follow PEP 8 style guidelines\n- Use type hints throughout the codebase\n- Maintain comprehensive test coverage\n- Document all public APIs with detailed docstrings\n- Use conventional commit messages\n- Leverage Pydantic for data validation and configuration\n- Implement proper async/await patterns for I/O operations\n\n## License\n\nThis project is licensed under the MIT License - see the [LICENSE](LICENSE) file for details.\n\n## Support\n\n- **Documentation**: Available in the `docs/` directory\n- **Issues**: [GitHub Issues](https://github.com/dwharve/pythonium/issues)\n- **Discussions**: [GitHub Discussions](https://github.com/dwharve/pythonium/discussions)\n\n## Acknowledgments\n\n- Official Model Context Protocol SDK and specification\n- The open-source Python community\n- Pydantic and FastAPI ecosystems for configuration and validation patterns\n- Contributors and maintainers\n\n---\n\n**Status**: Production-ready Beta - Core functionality stable, comprehensive test coverage, active development\n\n**Current Version**: 0.1.5 \n**Last Updated**: July 6, 2025 \n**Maintainer**: David Harvey\n",
"bugtrack_url": null,
"license": null,
"summary": "A modular MCP (Model Context Protocol) server for AI agents with plugin-based architecture",
"version": "0.1.5",
"project_urls": {
"Bug Reports": "https://github.com/dwharve/pythonium/issues",
"Documentation": "https://pythonium.readthedocs.io/",
"Homepage": "https://github.com/dwharve/pythonium",
"Source": "https://github.com/dwharve/pythonium"
},
"split_keywords": [
"mcp",
" model-context-protocol",
" ai-agents",
" plugin-framework",
" server",
" artificial-intelligence",
" tools"
],
"urls": [
{
"comment_text": null,
"digests": {
"blake2b_256": "5719e05e3270925a1ebb4364caea4d855620eacf80f0b9c222a3dd53b621fcc1",
"md5": "dadf4026ae4308aefec806262319d6de",
"sha256": "d4d0488f961b1822b1aced5402ccf64eda89bfb8c29742fae9a1e202a726381c"
},
"downloads": -1,
"filename": "pythonium-0.1.5-py3-none-any.whl",
"has_sig": false,
"md5_digest": "dadf4026ae4308aefec806262319d6de",
"packagetype": "bdist_wheel",
"python_version": "py3",
"requires_python": ">=3.8",
"size": 91293,
"upload_time": "2025-08-03T22:39:07",
"upload_time_iso_8601": "2025-08-03T22:39:07.151698Z",
"url": "https://files.pythonhosted.org/packages/57/19/e05e3270925a1ebb4364caea4d855620eacf80f0b9c222a3dd53b621fcc1/pythonium-0.1.5-py3-none-any.whl",
"yanked": false,
"yanked_reason": null
},
{
"comment_text": null,
"digests": {
"blake2b_256": "507f2ce42ddf85b5e20192e6a14ffc959ff3c1c689208b00637c021aefd93c40",
"md5": "60f4dafc73ccacbcaf1f2674effb9669",
"sha256": "c7360cfaa7332af4321b99a9e40f7eefa17670d04607b7c4983f3a47f1fc3677"
},
"downloads": -1,
"filename": "pythonium-0.1.5.tar.gz",
"has_sig": false,
"md5_digest": "60f4dafc73ccacbcaf1f2674effb9669",
"packagetype": "sdist",
"python_version": "source",
"requires_python": ">=3.8",
"size": 86451,
"upload_time": "2025-08-03T22:39:08",
"upload_time_iso_8601": "2025-08-03T22:39:08.965337Z",
"url": "https://files.pythonhosted.org/packages/50/7f/2ce42ddf85b5e20192e6a14ffc959ff3c1c689208b00637c021aefd93c40/pythonium-0.1.5.tar.gz",
"yanked": false,
"yanked_reason": null
}
],
"upload_time": "2025-08-03 22:39:08",
"github": true,
"gitlab": false,
"bitbucket": false,
"codeberg": false,
"github_user": "dwharve",
"github_project": "pythonium",
"travis_ci": false,
"coveralls": false,
"github_actions": false,
"requirements": [
{
"name": "pydantic",
"specs": [
[
">=",
"2.0.0"
]
]
},
{
"name": "pydantic-settings",
"specs": [
[
">=",
"2.0.0"
]
]
},
{
"name": "pyyaml",
"specs": [
[
">=",
"6.0"
]
]
},
{
"name": "click",
"specs": [
[
">=",
"8.0.0"
]
]
},
{
"name": "rich",
"specs": [
[
">=",
"13.0.0"
]
]
},
{
"name": "loguru",
"specs": [
[
">=",
"0.7.0"
]
]
},
{
"name": "httpx",
"specs": [
[
">=",
"0.24.0"
]
]
},
{
"name": "beautifulsoup4",
"specs": [
[
">=",
"4.12.0"
]
]
},
{
"name": "openai",
"specs": [
[
">=",
"1.0.0"
]
]
},
{
"name": "mcp",
"specs": [
[
">=",
"1.10.0"
]
]
},
{
"name": "langgraph",
"specs": [
[
">=",
"0.2.0"
]
]
},
{
"name": "orjson",
"specs": [
[
">=",
"3.9.0"
]
]
},
{
"name": "aiofiles",
"specs": [
[
">=",
"23.0.0"
]
]
}
],
"lcname": "pythonium"
}
JSON
