$ show wilsonfreitas/awesome-quant
This repository is a curated list (awesome-list) of open-source quantitative finance libraries, packages, and resources. It provides a structured README.md organized by category, with inline language tags, and a pipeline that parses the list, enriches it with GitHub metadata (stars, last commit dates), and generates a static website deployed to GitHub Pages. The repo also includes automated tooling for pull-request review: a validation script that checks entry format, section placement, URL security, duplicate detection, and a full CI-driven PR review workflow. The core logic resides in parse.py (multi-threaded GitHub API enrichment), site/generate.py (static site generator), scripts/validate_readme.py (entry validation), and scripts/review_pr.py (automated PR review with offline URL reachability checks and cross-PR duplicate detection). The project is primarily a reference list, not an application or data pipeline, but it includes high-quality supporting tooling for maintainers and contributors.
The LLM identified 2 files worth highlighting for reuse:
An automated pull-request reviewer for awesome-list contributions. Given a PR number, it fetches the diff, validates the added README entry against format rules, checks URL reachability (with a custom HTTPS client that avoids private/multicast addresses), detects duplicates within the base README and across open/recently-closed PRs, and verifies GitHub repository activity, documentation, and archive status. Well-tested with 30+ unit tests, and designed to run in CI (GitHub Actions) with security-conscious patterns (pinned connections, fail-closed on errors). Can be extracted and reused for any awesome-list repository with minimal changes.
A README entry validator that checks format, language tags, description period, HTTPS URLs, section placement, and duplicate names/URLs. Supports full validation and diff-based mode (validating only lines added relative to a git ref), making it suitable for both local and CI usage. Issues are categorized as errors or warnings, and duplicate detection uses normalized names and canonical URLs. The shared module scripts/readme_entries.py provides reusable parsing helpers (ENTRY_RE, extract_languages, slugify) that other scripts import.