Tools to scrape articles from a Substack publication and convert them into Markdown for a Hugo website.

• Python 3
• substack-api library

Recommended Setup (using venv):

python3 -m venv venv
source venv/bin/activate
pip install -r requirements.txt

Use scrape_substack.py to fetch an article by its URL. This saves the raw content and metadata to a JSON file in the scraped_data directory.

# Ensure venv is active or use direct path
./venv/bin/python3 scrape_substack.py <SUBSTACK_ARTICLE_URL>

Example:

./venv/bin/python3 scrape_substack.py https://{your_substack_url}.substack.com/p/{your_article_id}

Use process_letter_json.py to convert the scraped JSON files (in scraped_data) into Hugo-ready Markdown files (in output). this script handles cleaning HTML, removing widgets, and formatting the frontmatter.

./venv/bin/python3 process_letter_json.py scraped_data

The generated markdown files will be in the output/ directory and can be moved to your Hugo content content directory (e.g. content/posts/letters) as needed.

If you are using an AI agent and want to automate this flow, you can use the following prompt:

Task: Scrape and process this Substack article: [INSERT_URL_HERE]

Steps:

1. Run python3 scrape_substack.py [INSERT_URL_HERE] to fetch the data into scraped_data/.
2. Run python3 process_letter_json.py scraped_data to convert it to Markdown in output/.
3. Verify the generated markdown file in output/.