User repository on GitHub
The simplest way to write decision workflows in Python. Python with a smart if.
- Forks
- 0
- Last pushed
- Updated Sep 2026
Repo stats from the GitHub API, cached Oct 2026.
Jev directory
CommunityJev Repos & SDKs
Jevlang — sumanmichael
Python with a smart if: .jev files add the ~ question operator and jev/case blocks that Jev answers with a number or a label, never text. Files are rewritten to plain Python at import time; every ~ is one paid TypeSafe API call.
- Stars
- 5
- Language
- Python
- License
- MIT
- Reading time
- 3 min
- Published by
- Community
- Author
- sumanmichael
- Last verified
- Oct 2026
Highlights
- if ticket ~ "the customer wants a refund": fires at probability 0.5 or above; import jevlang installs an import hook so .jev files load like modules.
- No new interpreter: .jev files are rewritten to Python at import time, and a .py file shadows a same-named .jev file, so the hook cannot hijack a module.
- Callers receive float and string subclasses carrying .probs and friends, so the rest of your codebase never needs to know jevlang exists.
- python -m jevlang hello.jev runs a file; python -m jevlang --show prints the plain Python it becomes without making an API call.
- The README lists four things to know first: every ~ is a paid API call, the value left of ~ is what gets sent, and answers are probabilities, not facts.
Quickstart
bash
pip install jevlang
export TYPESAFE_API_KEY=...Watch out
Copy reviewed against the repo README on 2026-10-15; the captured text is truncated inside the agent-skill section. The author labels jevlang a prototype - tested and working, with limits listed in the README - and branches that matter need guards because answers are probabilities.
Questions about this entry
Short answers drawn from the linked project or page.
Two forms: x ~ question asks a question about a value and returns a number or a label, never text, and a jev/case block classifies into named cases. if ticket ~ "the customer wants a refund" is a probability comparison that fires at 0.5 or above.
No. .jev is Python plus two forms, rewritten to plain Python at import time. import jevlang installs an import hook so .jev files on sys.path import like ordinary modules, and a .py file always shadows a same-named .jev file, so the hook cannot hijack an existing module.
Python 3.10 or newer: pip install jevlang, export TYPESAFE_API_KEY, then python -m jevlang hello.jev runs a .jev file. python -m jevlang --show hello.jev prints the plain Python it becomes, without making an API call.
Every ~ is one call to TypeSafe's API that needs a key and is paid per call. The README's opening rules say the value left of ~ is what gets sent - so keep secrets out of it - and answers are probabilities, not facts, so guard the branches that matter.
Floats and strings - specifically subclasses carrying .probs and friends - so nothing else in your codebase needs to know jevlang exists. In the triage example, wants_refund(msg) returns a value usable directly in an ordinary if statement.
Yes - the README provides a copy-paste prompt with two installation paths: claude plugin marketplace add sumanmichael/jevlang plus claude plugin install in Claude Code, or npx skills add sumanmichael/jevlang for any other agent. Use one installation method.
The author labels it a prototype: tested, working, with its limits listed in the README. The four things to know stress that every ~ is a paid API call and that answers are probabilities rather than factual guarantees, so the README's own framing invites guarded use.
Other listings of this project
Forks and copies of this project in the directory. This entry is the canonical one.
A policy engine for LLM decisions: declare routes, gates and actions once in TypeScript or Python and every decision comes back validated, explainable, replayable and audited. The model answers only declared questions — application code owns the branch.
Jev Repos & SDKs#community#catalog-import#logicrw
Community
timmikeladze/jevlangJavaScript
timmikeladze/jevlangJavaScriptMore Jev resources like this
A CLI, Rust library and Python package that asks Jev typed Noul, Choice and Score questions through OpenRouter, then turns the probabilities into decisions with fuzzy rules — AND, OR, NOT and hedges into a support score per decision, or a crisp amount — and can draw the rule base as SVG.
Jev Repos & SDKs#community#catalog-import#logicrw
Community
dsaad68/fuzzy-jevRust
dsaad68/fuzzy-jevRustA policy engine for LLM decisions: declare routes, gates and actions once in TypeScript or Python and every decision comes back validated, explainable, replayable and audited. The model answers only declared questions — application code owns the branch.
Jev Repos & SDKs#community#catalog-import#logicrw
Communitytimmikeladze/jevlangJavaScript
timmikeladze/jevlangJavaScriptA Python library that turns Jev-style decision models into ordinary function calls: check returns a bool, classify picks one option, label picks every option that applies, and score rates on a scale, with probabilities behind every answer.
Jev Repos & SDKs#community#github#python
Community
jlowin/vibecheckPython
jlowin/vibecheckPythonA Bangla-focused adaptation of Laya, the non-autoregressive typed-decision model: targeted data preparation, fine-tuning and evaluation for classification, routing and scoring on Bangla, Banglish and code-switched text, with the checkpoint still to come.
Jev Repos & SDKs#community#github#python
Community
shamspias/laya-banglaPython
shamspias/laya-banglaPythonAn independent agent skill that walks a coding agent through designing a bounded Jev decision layer: contract, criteria, error behavior, fallback, and evaluation plan first, then scaffolding through TypeSafe or OpenRouter.
Jev Agents & MCP#community#catalog-import#logicrw
Community
grandamenium/jev-anythingPython
grandamenium/jev-anythingPythonA tuner that trains language models to answer closed-set decisions with a complete probability distribution in one forward pass, using a tokenized multi-class Brier loss so only the correct choice is needed to train.
Jev Repos & SDKs#community#catalog-import#madewithjev
Community
liushiliushi/jevtunerPython
liushiliushi/jevtunerPythonFrom the Jev guides
Original write-ups on what this entry is for.
From the Jev AI model community
Posts from builders shipping with Jev right now.