Enter repository
Paste a public GitHub repository URL. Deep links, .git suffixes and owner/repo shorthand all work.
GitHub repository analysis
Evaluate GitHub Projects in Seconds
Paste a repository URL and get a clear quality score out of 100. The analysis checks documentation, licensing, testing structure, project size, language and community signals — then shows exactly how every point was earned.
Deterministic scoringNo sign-upNothing storedPublic repositories
How it works
Three steps, no configuration
Analyze
The repository's metadata and root files are read through GitHub's public API and checked against eight fixed criteria.
Get your score
See the score, exactly which criteria passed and failed, and a list of concrete strengths and improvements.
What we evaluate
The complete scoring rubric
These are the only rules applied. Nothing is hidden, weighted dynamically, or inferred by a model — the same repository always scores the same.
DocumentationA README in the repository root (.md, .rst, .txt or no extension).
+20Project sizeMore than 100 KB of source — large enough to be a real project.
+20Primary languageGitHub can identify a primary programming language.
+15LicenseA LICENSE, LICENCE or COPYING file, or a license GitHub recognises.
+10Version control hygieneA .gitignore, so build output and secrets stay out of the repository.
+10TestingA tests/, test/, spec/ or __tests__/ directory, or a test config file.
+10Community interestAt least one star from another GitHub user.
+10CollaborationAt least one fork, indicating others have built on the work.
+5Maximum score100
Example result
What you get back
A score is never shown on its own. Every result lists each criterion, whether it passed, and the evidence found in the repository.
Portfolio score
75/ 100
Good Portfolio
The project shows solid implementation but could benefit from further improvements.
| Criteria | Result | Points |
|---|---|---|
| DocumentationFound README.md | Passed | +20 / 20 |
| Project size412 KB of source — substantial enough to review | Passed | +20 / 20 |
| Primary languageGitHub detected TypeScript | Passed | +15 / 15 |
| LicenseFound LICENSE | Passed | +10 / 10 |
| Version control hygieneFound .gitignore | Passed | +10 / 10 |
| TestingNo test directory or test config in the root | Missing | 0 / 10 |
| Community interestNo stars yet | Missing | 0 / 10 |
| CollaborationNo forks yet | Missing | 0 / 5 |
| Total | 5 of 8 met | 75 / 100 |
Sample output. Run a real analysis to see your own breakdown.
Why use it
Consistent review, in seconds
Fast
One URL, one request, a full result in a couple of seconds.
Consistent
Fixed rules and fixed weights, so every repository is judged the same way.
Actionable
Failed criteria come with the specific file or signal that was missing.
Transparent
The full rubric is published above — nothing about the score is a black box.
FAQ
Common questions
What repositories can I analyze?
Any public repository on github.com. Private repositories are not supported, and repositories on other hosts such as GitLab or Bitbucket are not analyzed.
Is my repository data stored?
No. Each analysis is a live read of GitHub's public API — nothing is written to a database, and there are no user accounts. The last five repositories you analyze are kept in your own browser's local storage so you can return to them, and that list never leaves your device.
How is the score calculated?
Eight deterministic checks with fixed point values, totalling 100. The same repository always produces the same score. There is no machine learning and no AI model involved — every point is traceable to a specific rule, and the breakdown on your result shows exactly which ones you earned.
Does this replace a technical interview?
No. It measures repository hygiene and project signals — documentation, licensing, testing structure, size and community activity. It does not read your code, assess architecture, or evaluate engineering ability. Treat it as a first-pass filter and a checklist, not a verdict on a candidate.
What does the score mean?
80 and above is a Strong Portfolio: the repository shows solid engineering practice. 60 to 79 is a Good Portfolio: a real project with room to improve. Below 60 means the repository needs more documentation, structure or supporting files before it presents well.
Why did my repository score lower than I expected?
The most common causes are a missing license file, no test directory in the repository root, or a repository under 100 KB. The breakdown lists every criterion with the exact evidence found, so you can see precisely what to add.