System One Adapter: compare typed decisions with LLMs

typesafe-ai/system-one-adapter-python

SDK / clientOfficial

System One Adapter runs Jev-shaped evaluations through OpenAI-compatible, Anthropic or Gemini APIs, preserving typed responses for model comparisons.

MIT · Python · updated 2026-09-22 · 356 ★ · Evidence checked Sep 30, 2026

What you can build with it

System One Adapter implements the typesafe_sdk evaluation interface using other providers’ LLM APIs.

It enables comparisons with TypeSafe while preserving typed answers and adding retry, latency and normalization diagnostics.

Configure a provider, model and probability or discrete answer mode; native structured outputs are optional.

Providers can omit token usage, which remains None. Corrective retries share an allowance within each evaluation.

Evidence checked: Sep 30, 2026

Repository description

Drop-in TypeSafeClient replacement backed by LLM APIs

Stars measure the whole repository, including work unrelated to Jev.

Official records the linked documentation or source. JevHunt has not independently run or benchmarked this project.

Evidence commit: e1d4cc938204b22fc5a3c3aca7044072fe3f712d

Discovered through: awesome-jev, typesafe-official.