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

jev-lint

Fuzzy linter hook for Claude Code and Codex where Jev checks each written file against team rules and reports violations at edit time.

Hookbeta
View on GitHub ↗Live demo ↗

Overview

jev-lint is a fuzzy linter for coding agents. When the agent writes a file, a hook asks Jev Noul questions about each applicable team rule in the .jev-lint/ folder and reports violations about 0.3 seconds later, so the agent fixes them before code review. It targets rules that need judgement rather than a regex, such as 'don't derive state in useEffect', 'a test must be able to fail' or 'never trust a client-sent userId'. Local thresholds decide what is reported, and API errors fail open.

The README reports that with the hook on, violations per task fell from 2.38 to 1.21 in Claude Code over 96 runs and from 1.67 to 0.92 in Codex over 48 runs, with method and results in an experiment notebook. It supports generating rules for a repository, learning from what it catches, rule packs with environment-variable options, optional Cloudflare Clef or OpenAI Decisions backends, local models, and an evaluation harness. A website is at jevlint.dev.

What the decision model decides

TypeQuestionAnswers
Yes / noDoes this file break this team rule?—

How it works

  1. Agent writes a file
  2. Hook selects applicable rules from .jev-lint/
  3. Jev answers a Noul per rule; local thresholds filter
  4. Report violations so the agent fixes them; fail open on errors

Strengths

  • Catches judgement-based rule violations early
  • Measured reduction in violations per task
  • Multiple backends including local models

Caveats

  • Demo video scene is illustrative
  • Rules need good wording and thresholds

Getting started

See the Install and 'Set it up with your agent' sections; website at https://jevlint.dev.

At a glance

Type
Hook
Status
beta
Language
TypeScript
Runtime
Claude Code and Codex hook
Has code
yes
Has tests
yes

Use cases and integrations

How curated lists describe it

  • Software engineering: uses Jev `Noul` judgments and local thresholds to flag team-rule violations as Claude Code and Codex edit, helping agents fix them before code review with configurable rule packs and repository-specific rules.

    yibie/awesome-jev

  • Unofficial fuzzy linter for Claude Code and Codex that uses Jev to flag team-rule violations at edit time, with configurable rule packs and repository-specific rules so agents can fix issues before code review

    AnotiaWang/awesome-decision-models

  • Asks Jev Noul questions about team-rule violations as Claude Code and Codex edit, using configurable rules and local thresholds to return feedback before code review; API errors fail open.

    cobanov/awesome-jev

  • A fuzzy linter for coding agents. It checks the code your agent writes against your team's best practices while the agent is still working, not at code review.

    hellogumbo/awesome-jev

  • Fuzzy linter for coding agents: a hook checks each file the agent writes against your team's rules in `.jev-lint/` with Jev and reports violations while the agent is still working.

    MrJev/awesome-jev

Frequently asked questions

What does Jev decide?

Whether each written file breaks each team rule.

Does it reduce violations?

The README reports roughly halving violations per task in Claude Code and Codex runs.

What if Jev is down?

The hook fails open.

Related projects

  • Abide

    Coding-agent hook that asks Jev one yes/no per project rule on every edit or turn and makes the agent fix any violation.

    HookbetaTypeScript
  • code-quality

    Advisor for coding agents: git and Stop hooks flag test tampering, secrets and skipped hooks, with optional Jev review findings.

    Hookbeta
  • Perch

    Semantic code linter that reads each method with its callers and callees and asks decision models about defects, kind, line and severity.

    CLIbetaTypeScript
  • jev-codes

    CLI and agent commands auditing a git diff against YAML standards packs, scoring each hunk with Jev so agents fix only flags.

    Agent skillbetaTypeScript
  • jeverifier

    Cheap Jev checks beside a Claude coding session: per-function code rules, doc contradiction checks and context retrieval.

    ProjectbetaPython
  • patdown

    Fuzzy linter CLI that checks a code tree against plain-language rules in one markdown file, judged by Jev via a swappable interface.

    ProjectbetaTypeScript