Skip to content

Lintus

Single-binary linter whose YAML rules are plain-language questions sent to the Jev model as yes-no judgments with probabilities, fitting CI jobs and pre-commit hooks.

Stars
12
Language
Rust
License
MIT
Published by
Community
Author
virolea
Last verified
Oct 2026

About this project

A linter whose rules are written in plain language. Lintus reads a YAML file of rules at the root of your repository. Each rule is a question about a file. For every file a rule applies to, Lintus asks the question to the Jev model as a noul (a true/false judgement with a probability), and reports an offense wherever the answer says so. It runs on the whole tree, on the files changed since a git ref, or on the files staged for a commit, so it fits a CI job as well as a pre-commit hook.

Quoted from the project's README. Read the full README.

Highlights

  • Rules declare question, criteria, paths, and thresholds, with offense flags for positively phrased checks.
  • Runs over the whole tree, named paths, uncommitted diffs, base-branch diffs, or staged content.
  • Text, GitHub Actions annotation, and JSON outputs with per-file probabilities for tuning.
  • Credential chain from flag to environment to saved login, with secrets recommended for CI.

Watch out

Each matching rule costs model judgment over full file contents, and phrasing decides what the probability means, so ambiguous questions need explicit criteria before tuning thresholds. 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.

More Jev resources like this

236GitHub stars
Rust governance kernel that sits between your coding agent and the repository, turning intent into bounded specifications, recorded state, and proof-backed publication across many separate invocations.
Jev Agents & MCP#community#jevbest
112GitHub stars
Rust CLI that searches a live source tree by describing what you want, returning files and line ranges with calibrated probabilities instead of building an index.
Jev Agents & MCP#community#jevbest
21GitHub stars
Rust CLI that sends selected source to the Jev API under JSON-defined choice rules and maps answers to warnings or errors.
Jev Agents & MCP#community#jevbest#catalog-import
35GitHub stars
Rust terminal tool that asks Jev noul, choice and score questions about piped text or files and returns calibrated probabilities with stable exit codes for shell scripts, CI jobs and agents.
Jev Agents & MCP#community#jevbest#catalog-import
64GitHub stars
LangGraph deep-research agent that plans, searches, reads pages, aggregates evidence, checks gaps, and writes a Markdown report with sources, built as a runnable tool and a staged learning reference.
Jev Agents & MCP#community#jevbest#catalog-import
14GitHub stars
Recursive language-model layer holding full inputs in a local evaluator and exposing a small working surface, with optional typed semantic judgments alongside generative calls.
Jev Agents & MCP#community#jevbest#catalog-import
Back to all resources

From the Jev community

Posts from builders shipping with Jev right now.

Follow @typesafeai