Most code quality tools tell you if your code is correct.
None of them tell you if it's efficient in the energy sense.
I wanted to change that. So I built sustainability-score —
a static-analysis tool that reads your repository and tells
you how its engineering choices affect carbon and cost
efficiency, grounded in the Green Software Foundation's
SCI specification.
The problem I was solving
The Green Software Foundation's Software Carbon Intensity
(SCI) specification gives us a framework for thinking about
software carbon. But measuring energy for a running system
is hard.
What isn't hard is reading a repo and asking: are the
engineering choices here pushing carbon up or down? That
is a static analysis problem. That is what this tool does.
How it works
The tool scores across five weighted pillars:
- Code / Algorithm Efficiency — 30%
- Cloud Infrastructure Choices — 25%
- Containerization — 15%
- CI/CD Practices — 15%
- SRE / Operations — 15%
Each pillar starts at 100 and loses points per finding
by severity. High: minus 20. Medium: minus 10. Low: minus 4.
The part I am most proud of — honesty
Every finding is stamped with a data quality tier:
- Tier 1 — static analysis only. Directional.
- Tier 2 — static plus declared infrastructure.
- Tier 3 — static plus operational telemetry.
- Tier 4 — direct measurement. Full SCI computable.
A static scan never claims above Tier 1. The report says
plainly it is directional, not a measurement. Carbon washing
is a real problem and I did not want to add to it.
Try it
pip install -e .
sustainability-score /path/to/repo --md report.md
A sample report is included in the repo so you can see
the output before running it.
What is next
- Broader per-language efficiency checks
- JSON schema for the output format
- A GitHub Action that posts advisory PR comments
- A proposal to the GSF reference implementations collection
MIT licensed. Open source. Contributions and feedback welcome.
GitHub: https://github.com/srinathgopinath-code/sustainability-score
Top comments (0)