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

Book of Answers

Chinese everyday-dilemma app where an LLM lays out the options and Jev picks one, revealed after you record your own leaning.

Projectbeta
View on GitHub ↗

Overview

Book of Answers (答案之书) is an AI app for everyday choices. You write down your dilemma and background; a large language model understands it, lists all reasonable and distinct candidate options (at least two, with no fixed upper limit) and writes explanations, and Jev picks one candidate and shows the probability distribution. Non-decision questions such as facts or chit-chat are answered directly without forcing options.

Before the reveal, you pick your own leaning or "hesitate"; Jev evaluates independently in the background and your choice is never sent to it. Results show both your pick and Jev's. Six reply styles change only the wording, never the candidates or probabilities. Answers can be saved, searched, re-read and followed up, with history kept in the browser. It supports TypeSafe and Vercel AI Gateway for Jev and needs an OpenAI-compatible LLM. Docs are in Chinese.

What the decision model decides

TypeQuestionAnswers
ChoiceWhich of the LLM-listed options should the user take?LLM-generated candidates

How it works

  1. User enters a dilemma
  2. LLM lists candidate options (or answers directly if not a decision)
  3. User records their own leaning
  4. Jev picks among candidates
  5. Show Jev's pick with probabilities and an explanation

Strengths

  • User's own choice kept separate from Jev's input
  • Style changes never alter the decision
  • Local browser history

Caveats

  • Needs both an LLM and a Jev key
  • Candidate quality depends on the LLM

Getting started

git clone https://github.com/csskrtao/jev-to-answer.git, set JEV_PROVIDER and JEV_API_KEY plus an LLM endpoint, then run with Node 22+.

At a glance

Type
Project
Status
beta
Runtime
Node web app
Has code
yes

Use cases and integrations

How curated lists describe it

  • This project integrates Jev to provide structured decisions for its workflow. See the repository for implementation details.

    logicrw/awesome-jev-projects

  • An LLM lays out the options for an everyday dilemma and Jev picks one, with a page for keeping and re-reading past answers. Chinese.

    MrJev/awesome-jev

Frequently asked questions

Who generates the options?

An LLM; Jev only picks among them.

Does my own choice affect Jev?

No, it is never sent to Jev.

Do reply styles change the answer?

No, only the wording.

Related projects

  • harden-jev-decides

    Decision board where one Jev request picks which of eleven livestream ideas becomes tonight's MVP, with code gating and weighting.

    DemoprototypeTypeScript
  • ask-jev

    Neo-brutalist Chinese web app styled after Ask Jeeves that gives direct verdicts on everyday dilemmas like whether to buy or what to eat.

    AppprototypeTypeScript
  • 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
  • a0-typesafe-ai

    Agent Zero plugin adding a typesafe_query tool that asks Jev Choice, Noul and Score questions and shows probability cards in chat.

    ProjectbetaPython
  • adk-go-typesafe

    Go integration bringing TypeSafe Choice, Score and Noul to Google's adk-go agents, with a typed client, routing agent and callbacks.

    ProjectbetaGo
  • dsh-jev

    DeepSeek Harness plugin adding a jev_evaluate tool for mixed Noul, Choice and Score evaluation of text or JSON.

    ProjectbetaTypeScript