mcp-agent and agentor
About mcp-agent
lastmile-ai/mcp-agent
Build effective agents using Model Context Protocol and simple workflow patterns
Fully implements MCP with automatic lifecycle management of server connections, and provides composable agent patterns (map-reduce, orchestrator, evaluator-optimizer, router) based on Anthropic's best practices. Scales from simple agents to production workflows via Temporal integration without API changes, enabling pause/resume/recovery capabilities. Supports the complete MCP specification including tools, resources, prompts, notifications, OAuth, and sampling.
About agentor
CelestoAI/agentor
Fastest way to build and deploy reliable AI agents, MCP tools and agent-to-agent. Deploy in a production ready serverless environment.
Provides a FastAPI-compatible MCP Server implementation (LiteMCP) with decorator-based tool definition and built-in authentication, plus a standardized Agent-to-Agent protocol using JSON-RPC messaging for multi-agent orchestration. Supports dynamic skill loading from Markdown files for context-aware task execution, and agents can be defined declaratively from Markdown with model/tool configuration.
Related comparisons
Scores updated daily from GitHub, PyPI, and npm data. How scores work