Agent Frameworks

LangGraph, Pydantic-AI, CrewAI, Mastra — frameworks for building data-pipeline agents.

🤖 5 skills listed 👥 Community ranked + Submit a skill
0
0↑ 0↓

AutoGen (Microsoft): Multi-Agent Conversations for Data Maintenance — converging into microsoft agent framework

by Microsoft
Visit →

AutoGen explained: Microsoft's multi-agent conversation framework, code-execution agents for data analysis, the AG2 fork, and AutoGen vs CrewAI vs LangGraph.

pip install autogen-agentchat autogen-ext[openai]
Python 3.10+ .NET OpenAI
0
0↑ 0↓

CrewAI multi-agent framework explained: role-based crews, Flows for deterministic pipelines, CrewAI vs LangGraph, and when multi-agent is overkill.

pip install crewai
Python 3.10+ OpenAI Anthropic
0
0↑ 0↓

LangGraph Stable

by LangChain
Visit →

LangGraph: low-level agent framework from LangChain for building stateful, multi-step data pipeline agents with explicit graphs.

pip install langgraph
Python 3.10+ LangChain OpenAI
0
0↑ 0↓

Pydantic-AI Stable

by Pydantic
Visit →

Pydantic-AI: type-safe agent framework from the Pydantic team. Structured outputs, function tools, and dependency injection for production data agents.

pip install pydantic-ai
Python 3.10+ OpenAI Anthropic
0
0↑ 0↓

smolagents explained: Hugging Face's minimal agent library where a CodeAgent writes Python instead of JSON tool calls — a great fit for data analysis work.

pip install smolagents
Python 3.10+ Hugging Face models OpenAI