User repository on GitHub
Self-hosted Jev decision workbench and Feishu bot: automatic choices, probabilities, and experimental word/character writing.
- Forks
- 0
- Last pushed
- Updated Sep 2026
Repo stats from the GitHub API, cached Oct 2026.
CommunityJev Practices & Patterns
Jev Bot — nssmd
Self-hosted Feishu bot plus web workbench where a text model proposes 2-5 candidate answers and TypeSafe Jev picks among them with probabilities.
- Stars
- 6
- Language
- JavaScript
- License
- MIT
- Published by
- Community
- Author
- nssmd
- Last verified
- Oct 2026
About this project
A self-hosted Feishu bot + web decision workbench powered by TypeSafe Jev. Ask a question, compare options and their probabilities, or experiment with writing by repeatedly choosing the next word or character. 中文部署说明 · TypeSafe API · License. Automatic decisions: a text model turns a question into 2–5 concrete alternatives; Jev independently chooses among them. Every candidate is shown and can be edited.
Quoted from the project's README. Read the full README.
Works forJev for Classification
Used inJev for Software & DevTools
- Languages
- JavaScript
Highlights
- Automatic mode drafts 2-5 editable candidates per question while Jev supplies the decision and probabilities.
- Manual mode takes 2-8 supplied options straight to Jev with no inserted fallback choice.
- Feishu covers private messages and group mentions with guided input, cancellation, and event deduplication.
- Web UI is dark and responsive with editable options, copyable results, and page-memory-only history.
Watch out
Self-host behind HTTPS with your own keys; the paid text backend and TypeSafe key are required for automatic decisions, and request text goes to both providers. Copy reviewed against the repository README on 2026-10-04, not from running it.
Questions about this entry
Short answers drawn from the linked project or page.
Yes. Supply 2-8 options yourself and Jev returns a probability distribution with no extra unknown option inserted. Automatic mode is only needed when you want the text model to draft the candidates.
The text backend must support the OpenAI Responses API with JSON-object output, configured via STRUCTURER_URL, STRUCTURER_MODEL, STRUCTURER_KEY, and STRUCTURER_ALLOWED_KEY. The key gate stops unrelated visitors from spending your text budget.
Backend and bot are separate processes needing Node.js 22.22+ with no frontend build. Browser keys and history clear on reload, the backend does not log credentials or prompts, and bot state keeps message IDs rather than bodies.
Treat them as an experiment: outputs may repeat, turn ungrammatical, or state false facts. They are a finite-vocabulary decision loop over letters and word buckets, not a general chat model.
It is a finite-vocabulary loop: word mode picks a function word or letter bucket then a complete word at default temperature 0.7, up to 24 words and 48 on the web, and character mode picks the highest-probability character from a-z and space up to 64 characters. There is no EOS or model-driven stop.
Requires Node.js 22.22+ and npm. git clone the repository, npm ci, copy .env.example to .env, chmod 600 .env, then npm start and open http://localhost:8787 to connect a TypeSafe key. There is no frontend build step.
The bot stores message IDs and processing states rather than message bodies, web history stays in page memory only, backend code does not log credentials, prompts or responses, and guided input is held in memory for up to 10 minutes.
Other listings of this project
Forks and copies of this project in the directory. This entry is the canonical one.
13GitHub stars
A market decision loop that feeds simulated stock, crypto and meme states to Jev or a local engine, then gates each answer. Execution is paper only and there is no live flag.
Jev Practices & Patterns#community#awesomejev#catalog-import
Community
bl888m/jev-botPython
bl888m/jev-botPythonMore Jev resources like this
23GitHub stars
Live race rig that labels 1,000 real app reviews with Jev typed questions against Gemini JSON output on identical batches, then replays recordings and re-ranks rows locally without extra model calls.
Jev Practices & Patterns#community#jevbest#catalog-import
10GitHub stars
Minecraft setup where Hermes plans at milestones, Jev picks one bounded harness action per step, and Mineflayer code owns validity and execution with no screenshots.
Jev Practices & Patterns#community#awesomejev#catalog-import
6GitHub stars
Watch Jev play chess against OpenRouter LLMs or Stockfish, or play Jev yourself, with every weighed move, likelihood, clock, and finished game saved for replay.
Jev Practices & Patterns#community#awesomejev#catalog-import
Community
choxos/jevchessJavaScript
choxos/jevchessJavaScript5GitHub stars
Inline verifier that checks LLM responses and agent plans with one Jev call across alignment heads and returns block, flag, or pass for gates and hooks.
Jev Practices & Patterns#community#awesomejev#catalog-import
Community
caiovicentino/jev-alignJavaScript
caiovicentino/jev-alignJavaScript4GitHub stars
OpenAI-compatible proxy that holds streamed tokens in a sliding buffer and scores the full criteria matrix before release, cutting the stream before a violating token reaches the client.
Jev Practices & Patterns#community#awesomejev#catalog-import
3GitHub stars
A single-file port of jev-ultrafast that lets Jev drive an Ego Lite browser on macOS: an in-page script indexes visible controls and one typed-choice request returns both the operation and the target index.
Jev Practices & Patterns#community#awesomejev#catalog-import
From the Jev guides
Original write-ups on what this entry is for.
Most Jev accuracy problems are state problems. What belongs in a state, what is a pre-made judgment to leave out, and why naming the slice you ask about matters more than the wording.
Jev Practices & Patterns#state#design#primitives
Read article
A Noul returns one number and no separate confidence. How to read the value, pin its boundary with criteria, split compound questions, and set a per-action threshold that matches the cost of being wrong.
Jev Practices & Patterns#noul#primitives#confidence
Read article
A Score is a weighted position, not a ruler. Why the number can land between levels, why levels are judged separately, and how to combine several atomic scores in code instead of one crowded rubric.
Jev Practices & Patterns#score#primitives#scoring
Read article
From the Jev community
Posts from builders shipping with Jev right now.