Langchain

• https://docs.langchain.com/oss/python/langchain/overview (you can copy it to input it to an LLM - see "llms.txt")
• LangChain is an open source framework with an agent architecture and integrations with several LLMs
• Build LLM-powered applications
• It abstracts the process of interacting with the LLM
• Supports agent ecosystem (e.g., search on internet, query database, send emails)

Documentation

• Reference https://reference.langchain.com/python/langchain/
• Docs https://docs.langchain.com/oss/python/langchain/overview

LangChain Expression Language (LCEL)

• It's new syntax introduced to simplify the creation of chains in LangChain applications

chain: Runnable = prompt_template | model
chain: Runnable = prompt_template.__or__(model) # same

Evolution of Agents

1. ReAct Prompt with AgentExecutor (2022~mid-2023)
   • create_react_agent & AgentExecutor
   • Prompt with the reasoning format (Thought/Action/Observation), tool descriptions and output instructions
   • The AgentExecutor ran the loop: call LLM, parse output, tool calls, append observations, repeat
2. Tool Calling with AgentExecutor (mid-2023~2024)
   • create_tool_calling_agent & AgentExecutor
   • OpenAI introduced function calling (June 2023)
   • Use the function calling functionality of LLMs (made available in June 2023) that also support the structured output
   • AgentExecutor still orchestrated the loop
3. Tool Calling with LangGraph
   • create_agent (langchain 1.0)
   • LangGraph allowed agents with graphs/state machines, not hidden loops