How birdger/daily-arxiv-digest works
This repository is a scheduled arXiv paper digest generator. GitHub Actions starts the script, which loads user-defined filters, fetches papers from RSS (falling back to the arXiv API), scores and filters them, then writes a Chinese Markdown digest and refreshes the README index. The script excerpt supports the configuration, retrieval, parsing, and scoring relationships; output and scheduling links are documented in the README. The workflow file itself was not sampled.
Main entry points: daily_digest.py Digest runner · daily.yml GitHub Actions · config.json Digest settings · digests Daily Markdown
Components 9
Automation
- Daily trigger
- GitHub Actions
.github/workflows/daily.yml
Digest pipeline
- Digest runner
scripts/daily_digest.py - Digest settings
config.json - Paper retrieval
scripts/daily_digest.py - Scoring and filtering
scripts/daily_digest.py
Published digests
- Daily Markdown
digests - README index
README.md
Other components
- arXiv feeds
Connections 8
- Daily trigger → GitHub Actions: triggers
README.md - GitHub Actions → Digest runner: runs daily
README.md - Digest runner → Digest settings: loads config
scripts/daily_digest.py - Digest runner → Paper retrieval: collects papers
scripts/daily_digest.py - Paper retrieval → arXiv feeds: fetches RSS/API
scripts/daily_digest.py - Digest runner → Scoring and filtering: scores papers
scripts/daily_digest.py - Digest runner → Daily Markdown: writes digest
README.md - Digest runner → README index: updates index
scripts/daily_digest.py