techXgood/techxgood
A platform to highlight innovative tech projects with altruistic goals, demonstrating how small efforts can lead to surprisingly impactful solutions for people or the environment. While humans endure for decades, code has the potential to last forever.
What's novel
A platform to highlight innovative tech projects with altruistic goals, demonstrating how small efforts can lead to surprisingly impactful solutions for people or the environment. While humans endure for decades, code has the potential to last forever.
Code Analysis
13 files read · 3 roundsA static website that crawls GitHub/GitLab for open-source projects with social/environmental impact, auto-categorizes them using sentence-transformer embeddings, and displays them as filterable cards with stats and charts.
Strengths
The NLP pipeline (embedding-based categorization, RAKE keyword extraction, extractive summarization) is a genuine implementation, not just a wrapper, and the GitHub Actions automation for continuous crawling and updating is well-structured. The Pydantic validation layer with provider-specific models (GitHub/GitLab) shows thoughtful data modeling.
Weaknesses
Zero tests, multiple bugs (broken env var in set_gh_token, HTTP side-effects in Pydantic validators, broken import path in add_single_project.py, workflow referencing a non-existent script path), and significant dead/commented-out code throughout. The 'database' is a single 1.7MB JSON file with O(n) lookups, and the frontend loads the entire file client-side with no pagination.
Score Breakdown
Signal breakdown
Innovation
Craft
Traction
Scope
Evidence
Commits
1890
Contributors
1
Files
62
Active weeks
52
Repository
Language
Python
Stars
2
Forks
0
License
GPL-3.0