I created a free Google Sheets tool that provides fair, nuanced evaluation of contributions.
The problem it solves:
GitHub: “Developer A has 1,913 additions” → appears to be top contributor
This tool: “Developer A: 1,810 pts (40% features, 32% bug fixes, 10% testing)” → shows actual work composition and value
Smart features: • Automatically excludes starter code and library imports • Skips merge commits (no credit for just combining branches) • Minimum line thresholds prevent gaming • Tracks net changes (editing same lines repeatedly doesn’t inflate score) • Uses GitHub username to prevent duplicate entries
Use cases: CS course grading, team analytics, recognizing diverse contribution types beyond just code
Free, open source, runs entirely in Google Sheets (no installation).
https://github.com/kyliemckinleydemo/github-contribution-tra...
KylieM•2mo ago
In my team projects, GitHub’s line counts created really misleading pictures of contribution. A teammate who built a critical authentication system (200 lines) looked like they contributed less than someone who copy-pasted 500 lines of boilerplate config files.
How it works: The tool analyzes commit messages, file types, and change patterns to categorize work into 8 types: • Feature Creation (3.0x default weight) • Bug Fixing (2.0x) • Refactoring (0.7x-0.1x based on scope) • Testing (0.4x) • Documentation (0.3x) • Support Code (0.5x)
You can customize these weights for your project. Building an MVP? Crank up features. Paying down tech debt? Boost refactoring.
Gaming prevention: Students definitely try to game these systems, so it includes: • Minimum 10-line threshold for code commits • Auto-detection of starter code (“provided files”, “libcs50”, etc.) • Merge commit exclusion • Net change tracking per file (editing same lines repeatedly doesn’t help) • Configurable start date to exclude instructor code
Real example: In one project, GitHub showed: • Team Member A: 1,913 additions (looked like top contributor) • Team Member B: 959 additions • Team Member C: 603 additions
This tool revealed: • Team Member B: Actually delivered most value (70% high-impact features + critical bug fixes) • Team Member A: Balanced work (good mix of features, bugs, testing) • Team Member C: Infrastructure specialist (build system, CI/CD - invisible in line counts)
Tech details: • Runs on Google Apps Script (no server needed) • Calls GitHub REST API directly • All data stays in your Google Sheet • Takes about 5 minutes to set up • Works with private repos (just needs a personal access token)
Limitations: • Single repo only (multi-repo aggregation on roadmap) • GitHub only (no GitLab/Bitbucket yet) • Can’t measure code quality or business impact (just categorizes work types) • Line count is still imperfect, but it’s objective and measurable
What’s next: Considering: • Time-series analysis (contribution trends over time) • Team comparison views • Export to PDF reports • GitLab support
Would love feedback on what would be most useful! Try it: https://github.com/kyliemckinleydemo/github-contribution-tra...