wehub-resource-sync
63 lines
2.8 KiB
Plaintext
63 lines
2.8 KiB
Plaintext
# ============================================================================
|
|
# A2A + AG-UI Starter - Python Agent Dependencies
|
|
# ============================================================================
|
|
|
|
# ============================================================================
|
|
# AG-UI Protocol Implementation
|
|
# ============================================================================
|
|
# ag-ui-adk: Allows ADK agents to communicate with the frontend via AG-UI Protocol
|
|
# This is used by the orchestrator agent to receive messages from the UI
|
|
ag-ui-adk>=0.0.1
|
|
|
|
# ============================================================================
|
|
# A2A Protocol Implementation
|
|
# ============================================================================
|
|
# a2a: Core A2A Protocol SDK for agent-to-agent communication
|
|
# a2a-sdk: Additional utilities including HTTP server support
|
|
# These packages enable the Research and Analysis agents to communicate via A2A
|
|
a2a>=0.1.0
|
|
a2a-sdk[http-server]
|
|
|
|
# ============================================================================
|
|
# Google ADK (Agent Development Kit)
|
|
# ============================================================================
|
|
# google-adk: Google's framework for building AI agents with Gemini models
|
|
# Used by: Orchestrator Agent, Analysis Agent
|
|
# litellm: Unified interface for multiple LLM providers (dependency of ADK)
|
|
google-adk>=0.1.0
|
|
litellm>=1.0.0
|
|
|
|
# ============================================================================
|
|
# LangGraph Framework
|
|
# ============================================================================
|
|
# langgraph: LangChain's framework for building stateful agent workflows
|
|
# langchain: Core LangChain library for LLM applications
|
|
# langchain-openai: OpenAI integration for LangChain
|
|
# Used by: Research Agent
|
|
langgraph>=0.2.0
|
|
langchain>=0.3.0
|
|
langchain-openai>=0.2.0
|
|
|
|
# ============================================================================
|
|
# Web Server
|
|
# ============================================================================
|
|
# fastapi: Modern Python web framework for building agent HTTP endpoints
|
|
# uvicorn: ASGI server for running FastAPI applications
|
|
# All agents expose HTTP endpoints using FastAPI
|
|
fastapi>=0.115.0
|
|
uvicorn>=0.30.0
|
|
|
|
# ============================================================================
|
|
# Utilities
|
|
# ============================================================================
|
|
# python-dotenv: Loads environment variables from .env file
|
|
# Used to configure API keys (GOOGLE_API_KEY, OPENAI_API_KEY)
|
|
python-dotenv>=1.0.0
|
|
|
|
# ============================================================================
|
|
# LLM Providers
|
|
# ============================================================================
|
|
# openai: OpenAI Python SDK for GPT models
|
|
# Used by the Research Agent (LangGraph)
|
|
openai>=1.0.0
|