swift-jev: A Swift client for TypeSafe Jev
A Swift library and CLI for TypeSafe AI's Jev System One model, offering typed questions and calibrated probability answers.
OPEN SOURCE · CURATED REPOSITORIES
Curated Jev SDKs, tools, demos, integrations, benchmarks, and research projects. Repository metrics come from GitHub; titles, summaries, and categories are editorial.
Sponsors
These products pay for the directory. Want yours up here?
Ranked by GitHub Stars
Ranked by recent Star growth and update activity
A Swift library and CLI for TypeSafe AI's Jev System One model, offering typed questions and calibrated probability answers.
A TypeScript starter demonstrating confidence routing, fallbacks, evaluation, and RAG patterns built on the official TypeSafe AI Jev SDK.
kojev is an unofficial Kotlin Multiplatform client for TypeSafe AI's Jev/System One model, returning enum or sealed types instead of string keys.
TypeSafe Jev-powered OpenCode plugin that detects ambiguous user requests and asks clarifying questions before the agent acts.
Checks your code against your Markdown specs on every commit, using TypeSafe AI's Jev model to return calibrated probabilities against thresholds.
One card in the feed, every sixteen, on every page of it. $50 a week, cancel any week.
Implements a Jev-style /v1/systemone API on top of any OpenAI-compatible backend, using logprobs and a restricted softmax for typed answers with probabilities; includes measured performance and model-specific notes.
OpenJev is an open-source System One model implementation compatible with TypeSafe AI's Jev. It returns calibrated probability distributions without text generation and works with the official typesafe-sdk unchanged.
Scores every test against a git diff with TypeSafe's Jev/System One model and emits filter arguments for vitest, node:test, Playwright, cargo test, and go test.
A Claude Code plugin that scores review findings, debug hypotheses, and design options with TypeSafe Jev, returning calibrated probabilities instead of one more opinion.
An OpenCode plugin that uses TypeSafe Jev to judge and remove stale tool calls/results from outgoing model requests while keeping the session log intact.
A Claude Code plugin that calls TypeSafe Jev to rank installed skills per prompt, logs decisions in shadow mode, and documents why it is unlikely to help a strong model.
Wraps any CLI and uses TypeSafe's Jev to suggest correct subcommands by intent, with offline fallback.