DEV Community

Samantha Start
Samantha Start

Posted on

I Scanned 10 Popular TypeScript Repos — Stars Don't Predict Production Readiness

I couldn't help myself. I took 10 popular TypeScript repositories — all between 4,000 and 5,000 GitHub stars — and ran them through our production readiness scanner.

The results surprised me.

The Scores

Repo Stars Score What I noticed
mengxi-ream/read-frog 4,941 79/100 Translation browser extension with its infrastructure together. Genuinely impressed.
tagspaces/tagspaces 4,985 77/100 Offline-first document manager. When your app works offline, you have to get the engineering right.
useplunk/plunk 4,933 75/100 Open-source email platform. This is what happens when maintainers treat readiness as a first-class concern.
microsoft/FluidFramework 4,917 72/100 Real-time collab library. Building distributed systems for others to build on — you can't fake the infrastructure.
extension-js/extension.js 4,963 66/100 Cross-browser extension framework. Solid. Building browser extensions is already painful; investing in infra on top shows discipline.
microsoft/tsdoc 4,943 65/100 TypeScript doc comment standard. Expected higher from Microsoft tbh — but it's a spec project, not an app.
opennextjs/opennextjs-aws 4,974 60/100 Next.js adapter for AWS. Someone thought about deployment beyond "it works on my machine."
gluestack/gluestack-ui 4,994 49/100 React/React Native component library. Nearly 5K stars, below 50. The components are gorgeous. The infrastructure has gaps.
tinyplex/tinybase 4,980 44/100 Reactive data store. Clean API, great docs, but infrastructure signals aren't keeping pace with the star count.

What I Learned

Stars don't predict production readiness. At all.

The highest-starred repo in my set (gluestack-ui, 4,994 stars) scored 49. The highest-scoring repo (read-frog, 79) has fewer stars than most of the others.

The repos that scored well had something in common: someone invested in CI, tests, dependency management, and configuration as foundational work, not as an afterthought.

The repos that scored poorly had great code and great docs — but the engineering infrastructure around the code was thin or missing.

The Pattern

Every time I scan repos, the same pattern emerges:

  • The code is usually fine. Whether human-written or AI-assisted, the application logic works.
  • The infrastructure is where it breaks. CI pipelines, test coverage, branch protection, secrets management, dependency health — these are the signals that separate "it runs" from "it's production-ready."
  • Stars measure popularity, not readiness. A repo can have 5,000 stars and still be missing half the production basics.

What We Check

Our scanner looks at 9 production readiness signals:

  1. CI enforcement
  2. Test coverage
  3. Type safety
  4. Dependency health
  5. Branch protection
  6. Dead code
  7. Dead exports
  8. Linter configuration
  9. Route coverage

Each signal is a binary or scored check. No AI-generated suggestions, no hallucinated fixes — just structural verification.

Try It

If you maintain an open-source TypeScript project (or any project, really), I'm genuinely curious how it scores. The scanner is free, no signup required:

repofortify.com

Paste your repo URL, get a score in seconds. I'd love to see what the dev.to community's projects look like.

And if you score above 75 — seriously, tell me. I want to celebrate the repos that are doing it right.


Samantha Start builds production readiness scanning tools at RepoFortify. She scans too many repos and can't stop talking about what she finds.

Top comments (0)