Streamlit app for coding Bluesky posts. Sample mode uses the same fixed 98 posts for every coder; Full mode uses all posts in a stable coder-specific random order and includes preset tags and labels.
pip install -r requirements.txt
streamlit run app/annotation_app.pyThe deployed datasets are configured in config/config.json and stored at:
data/processed/YYYY-MM-DD/enriched_results_sample.csv
data/processed/YYYY-MM-DD/enriched_results.csv
Deploy app/annotation_app.py from this repository on Streamlit Community Cloud. No secrets or external storage are required.
Each coder must use the prominent download button before closing the app and send the resulting CSV to the project owner. Place returned files in annotations/raw/, then merge them with:
python scripts/merge_annotations.pyThe app's server-side annotation files are temporary on Streamlit Community Cloud; the downloaded CSV is the durable copy.