How ready is your SDK for AI coding agents?
When a library ships a breaking major, AI assistants keep writing the old API — it looks right, but it doesn't compile against the new version. SDKProof measures exactly how much: it has a model solve real tasks, then type-checks the code against your real installed package.
Scorecards
Each score is claude-opus-5 across ~10–15 realistic tasks, type-checked against the current package.
Readiness tracks the drift window
The score isn't about how smart the model is — it's about how recently your SDK changed. TanStack Query v5 (2023) is fully absorbed — a perfect 100. The 2025 majors (Zod, the Vercel AI SDK) are now fully absorbed too, at 100 — Opus 5 closed the gaps Opus 4.8 still had. The freshest releases are where the agent still writes the old API: Next.js 16 misses only its brand-new 2-arg revalidateTag() (92), while Prisma 7's driver-adapter overhaul keeps it at 87.
Right after a major the gap is wide; as models retrain it narrows — then re-opens on your next major and shifts on every model release. It's a moving number, which is why it's worth monitoring, not auditing once.
Three steps, one honest signal
Generate
A model solves ~10–15 realistic tasks for the library. Prompts name the functions, never the option names.
Type-check
Each solution is dropped into a project with the real installed package and run through tsc --noEmit.
Score
Pass = compiles clean. Failures are the compiler's own diagnostics — hallucinated members, wrong args, removed options.
Scoring soon
@remix-run/* imports removed
Clerk v6 — auth() now async
Stripe — resources shift across API versions
Score my SDK
Name any TypeScript package & I'll type-check how well AI writes its current API — or run it yourself, it's open source.