GitHub ↗

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.

// no LLM judge — the compiler decides. pass = compiles clean.
The board

Scorecards

Each score is claude-opus-4-8 across ~10–15 realistic tasks, type-checked against the current package.

Prisma 7
@prisma/client
80/100
Nails the query API, but still writes removed v6 setup — new PrismaClient(), datasources, $use.
View scorecard →
Vercel AI SDK 7
ai
90/100
Mostly caught up, but wires tools the old way — parameters (now inputSchema) and removed maxSteps.
View scorecard →
Zod 4
zod
90/100
Even nails the new 2-arg z.record() — only slips on the removed required_error option.
View scorecard →
What the scores mean

Readiness tracks the drift window

The score isn't about how smart the model is — it's about how recently your SDK changed. Zod 4 and the Vercel AI SDK shipped their breaking changes in 2025, so today's models have largely absorbed them (90 each). Prisma 7 is newer, and the same model still writes its old v6 setup (80).

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.

How it works

Three steps, one honest signal

01

Generate

A model solves ~10–15 realistic tasks for the library. Prompts name the functions, never the option names.

02

Type-check

Each solution is dropped into a project with the real installed package and run through tsc --noEmit.

03

Score

Pass = compiles clean. Failures are the compiler's own diagnostics — hallucinated members, wrong args, removed options.

On the board next

Scoring soon

Score your own SDK

Point SDKProof at any TypeScript package and see what agents get wrong. It's open source — run it in a few minutes.

Get it on GitHub →
early project · scorecards are independent analysis · the compiler is the judge