Skip to content
Metatext Decide — typed decisions for your agents, with a probability on every answerGet early access
Metatext
Sign Up

rust-sysone

Very early unofficial Rust client for TypeSafe System One with a fluent request builder and typed question and answer enums.

LibraryprototypeApache-2.0 OR MIT
View on GitHub ↗

Overview

rust-sysone (crate sysone) is an unofficial Rust client for TypeSafe AI's System One API by Jeremy Chone, author of the genai crate, and part of the zcoder.run libraries. It is an early 0.0.x release with a deliberately basic API surface.

Requests use a fluent builder: Request::from_state takes any JSON-convertible state, and questions are appended by named or indexed keys using Question::noul, Question::choice with criteria, or score questions with ordered levels. Responses report the model, input and output tokens, optional cost, and typed answers (Noul with a probability, Choice with choice, confidence and probabilities, Score with score, confidence, probabilities and legend). The key comes from TYPESAFE_API_KEY or the builder. It is dual-licensed Apache-2.0 or MIT.

What the decision model decides

TypeQuestionAnswers
Yes / noDoes it compile?—
ChoiceClassify code statusok · fix
ScoreRate severityLow · High

How it works

  1. Create a Client (key from env or builder)
  2. Build a Request from state and append questions
  3. client.exec(req).await
  4. Read typed answers, tokens and cost

Strengths

  • Simple fluent API
  • Reports cost per call

Caveats

  • Very early 0.0.x release; minimal functionality
  • Unofficial

Getting started

Add the sysone crate, set TYPESAFE_API_KEY, then Client::default().exec(Request::from_state(...).append_question(...)).

At a glance

Type
Library
Status
prototype
Language
Rust
Runtime
Rust library (tokio)
Has code
yes
License
Apache-2.0 OR MIT

Use cases and integrations

How curated lists describe it

Frequently asked questions

Is it production-ready?

No; the README calls it a very early release with basic functionality.

Who maintains it?

Jeremy Chone, author of the genai crate, as part of zcoder.run.

Does it report cost?

Yes, responses include tokens and an optional cost.

Related projects

  • ai

    Hack Club's lightweight LLM proxy, listed for adding a Jev forwarding endpoint behind its auth, limits and usage logging.

    MCP serverproductionTypeScript
  • Ax

    Ax, a DSPy-style multi-language LLM framework, listed with a native TypeSafe provider for Boolean, Choice, Score and raw Jev questions.

    PluginproductionTypeScript
  • Effect

    Effect TypeScript library whose @effect/ai-typesafe package maps classify, probability and rating operations to Jev questions.

    ProjectproductionTypeScript
  • instructor-php

    PHP monorepo whose Polyglot module adds a TypeSafe driver for typed Noul, Choice and Score decisions alongside LLM inference.

    ProjectproductionPHP
  • jev-mcp (freepik-company)

    Go MCP server giving any client typed Jev decisions via OpenRouter or TypeSafe, with validated answers and no paid retries.

    MCP serverproductionGo
  • llm-typesafe

    Simon Willison's LLM plugin that adds Jev as a model for yes/no, choice and score questions from the command line.

    PluginproductionPython