An application for testing docs procedures using Langflow.
TODO: Clean up the output in the frontend app.
Create a web application that sends data to Langflow and returns a useful result. Export a flow from the IDE, then serve the flow in headless mode in Docker. Post to the served endpoint from a frontend, run the flow, get a result.
- Build image:
docker build -t solvedcaptcha/langflow-pokedex-agent:1.0.0 .
- Run container:
docker run -p 7860:7860 solvedcaptcha/langflow-pokedex-agent:1.0.0
- Start frontend application and ask about the Pokemon you're interested in.
cd frontend
npm install
npm run dev
You should get back a full page article and a picture, generated by the Poke API using Langflow. Right now it just returns the event stream.
You can drop any flow into the /flows folder and build the Docker image from root. You don't need the toml and the uv.lock unless you're running locally.
- Get the flow-id for a flow you loaded (top of the flow file under
id) and execute a curl against it to confirm it's being served. - Add this value to the frontend app.
If you prefer to run the application locally without Docker, follow these steps:
- Create a Python virtual environment (Python 3.10-3.13 required):
python -m venv .venv
source .venv/bin/activate # On Windows, use `.venv\Scripts\activate`- Install the project dependencies:
pip install -e .- Set up your environment variables:
- Copy the
.env.examplefile to.env(if it exists) - Add your required API keys and configurations
- Start the Langflow server:
langflow run --env-file .env- In a separate terminal, start the frontend:
cd frontend
npm install
npm run devThe Langflow server will be available at http://localhost:7860, and the frontend will be at http://localhost:3000.
Note: Running locally requires managing Python dependencies and Node.js packages on your machine. Using Docker is the recommended approach for consistency across environments.