README.md

LangChain Agents with LangSmith

This streamlit walkthrough shows how to instrument a LangChain agent with tracing and feedback. It highlights the following functionality:

• Implementing an agent with a web search tool (Duck Duck Go)
• Capturing explicit user feedback in LangSmith
• Linking to the run trace for debugging

Below is an example:

Prerequisites

The requirements for this streamlit application are listed in the requirements.txt file.

(Recommended) First, create and activate virtual environment.

python -m pip install -U virtualenv pip
python -m virtualenv .venv
. .venv/bin/activate

Then install the app requirements.

python -m pip install -r requirements.txt

Next, configure your API keys for LangSmith and the LLM provider (we are using OpenAI here for the LLM).

export OPENAI_API_KEY=your-openai-api-key
export LANGCHAIN_TRACING_V2=true
export LANGCHAIN_API_KEY=your-langsmith-api-key
export LANGCHAIN_PROJECT=langsmith-streamlit-agent

Finally, start the streamlit application.

python -m streamlit run main.py

You can interact with it, leave feedback, and view the traces to see what's going on under the hood.