dedalus-mcp-python and mcp-server
One tool is a Python framework for creating MCP models, while the other provides a server for deploying and managing those models, making them complementary components within the MCP ecosystem.
About dedalus-mcp-python
dedalus-labs/dedalus-mcp-python
A simple and performant Model Context Protocol framework for Python.
Wraps the official MCP SDK with decorator-based registration (decoupling functions from specific server instances), implements strict protocol versioning with capability dataclasses per spec version, and validates responses against JSON schemas to catch structural drift. Built for teams with existing infrastructure—no bundled auth, CLI scaffolding, or middleware opinions—and ships at 122 KB with every feature traced to MCP spec clauses for debuggability.
About mcp-server
keboola/mcp-server
Model Context Protocol (MCP) Server for the Keboola Platform
Exposes Keboola's data pipelines, transformations, and job orchestration as composable tools via the Model Context Protocol, enabling AI agents to query tables, create SQL transformations, trigger jobs, and manage development branches through a unified interface. Supports both a hosted remote server with OAuth authentication (for Cursor, Claude, Windsurf, and other MCP clients) and local deployment with configurable tool authorization via HTTP headers for access control. Built on Streamable HTTP transport with granular read-only and tool-filtering capabilities to safely integrate Keboola workflows into AI agent workflows.
Related comparisons
Scores updated daily from GitHub, PyPI, and npm data. How scores work