Measure the health of your open source project — and know exactly what to fix.
Percentile-based scoring across 5 dimensions. Calibrated against 2,400+ real GitHub repositories.
RepoPulse analyzes GitHub repositories and produces a composite OSS Health Score — a percentile-based rating computed from five dimensions, each weighted by importance. You get a clear picture of where a project stands and actionable recommendations to improve it.
Try it now — paste any GitHub repo URL into repopulse-arun-gupta.vercel.app and get a health report in seconds.
No account? No problem. The
/demoroute ships with six pre-analyzed repositories and one organization inventory — full scorecards, comparisons, and recommendations, no OAuth required. Fixtures refresh weekly.
RepoPulse is built for open source projects across the full maturity spectrum. The primary audience is:
- Maintainers of any open source project looking to understand and improve its health
- OSPOs and program managers evaluating OSS health across a portfolio
- Foundation-track projects (CNCF, Apache, Linux Foundation, etc.) preparing for stage reviews
Solo and early-stage projects are supported. RepoPulse auto-detects solo projects (≤2 contributors or maintainers) and switches to a re-weighted scoring profile that emphasizes Activity, Security, and Documentation — dropping Contributors and Responsiveness, which aren't meaningful signals at that scale.
The app is organized around three modes, each accessible from the top tab bar and directly deep-linkable:
| Mode | URL | What it does |
|---|---|---|
| Repos | /?mode=repos |
Analyze one repo or compare up to 4 side-by-side |
| Org | /?mode=org |
Browse all repos in a GitHub org with an org-level health summary |
| Foundation | /?mode=foundation |
Run foundation-track readiness scans (CNCF, Apache) |
Every view is shareable — the selected tab, org, foundation track, and input are all encoded in the URL. Use the Copy Link button in the tab bar to grab the current URL at any point.
| Feature | Description | |
|---|---|---|
| 📊 | OSS Health Score | Composite percentile across 5 scored dimensions, auto-adapted for solo or community projects |
| 📈 | Percentile Scoring | Every metric shows where the repo ranks vs. 2,400+ calibrated repos |
| 👥 | Repo Comparison | Side-by-side analysis of up to 4 repositories |
| 🏢 | Org Inventory | Org-level health summary across all scored dimensions plus Governance; repo table supports structured search prefixes (lang:go, stars:>500, archived:false, topic:kubernetes) |
| 🌱 | Foundation Track | CNCF Sandbox readiness scan (0–100 score, TAG recommendation, landscape detection); org-wide candidacy scan ranks every repo by readiness |
| 💡 | Recommendations | Actionable improvement suggestions across all scoring dimensions |
| 💬 | AI Chat | Ask natural-language questions about any analysis; 5 free chats/day per GitHub login, or bring your own API key |
| 📤 | Export | JSON, Markdown, and Copy Link from any view |
| 📖 | Scoring Methodology | Full transparency into calibration data and thresholds at /baseline |
RepoPulse scores each dimension as a percentile relative to repos in the same star bracket (Emerging, Growing, Established, Popular). The weighted composite becomes the overall health score.
Community profile (3+ contributors — default for most projects):
| Dimension | Weight | What it evaluates |
|---|---|---|
| Activity | 25% | PR throughput, issue flow, commit cadence, release frequency |
| Responsiveness | 25% | Issue/PR response times, resolution speed, backlog health |
| Contributors | 23% | Contributor concentration, repeat/new contributor ratios |
| Security | 15% | OpenSSF Scorecard, dependency automation, branch protection |
| Documentation | 12% | README, CONTRIBUTING, LICENSE, CODE_OF_CONDUCT, licensing & compliance, inclusive naming |
Solo profile (auto-detected for ≤2 contributors — Contributors and Responsiveness are not meaningful signals at this scale):
| Dimension | Weight | What it evaluates |
|---|---|---|
| Security | 35% | OpenSSF Scorecard, dependency automation, branch protection |
| Documentation | 35% | README, CONTRIBUTING, LICENSE, CODE_OF_CONDUCT, licensing & compliance, inclusive naming |
| Activity | 30% | PR throughput, issue flow, commit cadence, release frequency |
The app also surfaces a qualitative Governance analysis (2FA enforcement, branch protection, CODEOWNERS, maintainer bus-factor) and a Release Health breakdown that do not feed into the composite score but are shown as standalone tabs.
# 1. Clone and install
git clone https://github.com/arun-gupta/repo-pulse.git
cd repo-pulse
npm install
# 2. Set up environment (create a GitHub OAuth App first)
# Callback URL: http://localhost:3000/api/auth/callback
echo "GITHUB_CLIENT_ID=your_id" >> .env.local
echo "GITHUB_CLIENT_SECRET=your_secret" >> .env.local
# Optional: enable server-side AI Chat (any one key enables 5 free chats/day per GitHub login)
# echo "ANTHROPIC_API_KEY=sk-ant-..." >> .env.local
# echo "OPENAI_API_KEY=sk-..." >> .env.local
# 3. Run
npm run devOpen http://localhost:3000 and analyze your first repo.
See docs/DEPLOYMENT.md for Vercel deployment instructions.
| Phase | Focus | Status |
|---|---|---|
| 1 | Web app + core scoring (Activity, Responsiveness, Contributors, Documentation) | ✅ Done |
| 2 | Expanded scoring — Security, Licensing & Compliance, Inclusive Naming, Governance, Community, Release Health, Development Cadence, Project Maturity, Org Governance Audit, CNCF Sandbox Readiness | ✅ Mostly done (Foundation-aware recommendations for Apache/LF pending) |
| 3 | Integrations (GitHub Action, MCP Server, CLI, PR bot, VS Code, Badge, Webhook) | Planned |
| 4 | Git provider support (GitLab, Bitbucket, Gitea) | Planned |
See docs/PRODUCT.md for detailed feature specifications per phase.
We welcome contributions! See CONTRIBUTING.md for setup instructions, development workflow, and how to submit a pull request. Check out issues labeled good first issue to get started.
docs/PRODUCT.md— Product definition and feature specificationsdocs/DEVELOPMENT.md— Development workflow, implementation order, verification commandsdocs/DEPLOYMENT.md— Vercel deployment guidedocs/scoring-and-calibration.md— Scoring methodology, calibration pipeline, statistical approachdocs/ai-chat.md— AI Chat feature: providers, free tier, bring-your-own-key setup