This helper is used to generate the analytics digest twitter thread based on the Dune dashboard.
The script fetches data from the Dune Dashboard, processes it, and generates a markdown file that serves as the basis for the Twitter thread. The markdown file is then saved in the threads directory with the date as the filename.
The script also generates graphs based on the fetched data and saves them in the graphs directory.
The script uses the dune-client package to fetch data from the Dune Dashboard. Dune data is processed in chunks for each query, before being sent to GPT3.5/4 to write each section of the thread and then compose the entire thread.
The script also uses the matplotlib and seaborn packages to generate relevant graphs based on the fetched data.
- Clone the repository.
- Install the required Python packages using the command
poetry install. - Set up the environment variables. Example environment variables in
.env.example - Run the script using the command
python3 main.py. The script can also be run with specified start and end dates withpython3 main.py --sd <START_DATE like "2023-07-31"> --ed <END_DATE like "2023-08-06">to specify the digest dates manually.