User repository on GitHub
Configurable semantic linting powered by Jev, with file-level NOUL judgments and a magic-strings plugin.
- Forks
- 0
- Last pushed
- Updated Sep 2026
Repo stats from the GitHub API, cached Oct 2026.
CommunityJev Agents & MCP
JevLint — huntedman
Semantic lint that checks code against plain-English conventions via Jev and reports file-level violation probabilities for agents and CI.
- Stars
- 12
- Language
- TypeScript
- License
- MIT
- Published by
- Community
- Author
- huntedman
- Last verified
- Oct 2026
This is one of several Jev listings for the same project. The canonical entry is JevLint — iamtoomas.
About this project
Write coding conventions in plain English. JevLint checks what your code means, semantically, so your coding agent can fix what syntax rules miss. Use it in a write → check → fix → repeat workflow: your agent writes code, JevLint checks it against your conventions, and the agent uses the findings to make the next edit. Run it from your terminal, an agent’s verification step, or CI. Your agent or workflow controls when to run checks and how to fix findings; JevLint supplies the judgments.
Quoted from the project's README. Read the full README.
Works forJev for Classification
Used inJev for Software & DevTools
- Languages
- TypeScript
Highlights
- Built-ins cover magic strings and descriptive names, each scopable to separate files with its own ignores.
- Custom rules export a plugin object with a yes/no instructions question and a fix message from a folder or file.
- JSON output, threshold flag, dry-run request preview, exit codes 0, 1, and 2, plus pretty or compact text.
- Installs as a dev dependency on Node 24 or newer, loads JEV_API_KEY from local env, and never overwrites config.
Watch out
Findings lack line diagnostics and auto-fixes, dry runs still execute plugin code with full Node permissions, and secrets exclusions are filename rules rather than value detection. 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.
It judges what code means against described conventions so agents catch vague names and symbolic strings that syntax rules miss, then agents decide edits in a write-check-fix loop.
It excludes secrets and credentials directories and filenames with supported extensions, but that is filename matching rather than value detection, so other files can still send sensitive values.
The default 0.8 threshold flags file-level probabilities that are review signals rather than proof, with JSON reports carrying model, threshold, results, and summary plus skipped-file notes.
No. Dry-run skips the API and key but still imports and runs configured JavaScript and TypeScript plugins, which can read, modify, or reach the network, so run only trusted configs.
Node 24 or newer, then npm install --save-dev @jevlint/cli and npx jevlint init. Init writes jevlint.config.json with magic-strings enabled, prettyPrint on, and exclusions for tests and TypeScript declaration files, and it never overwrites an existing config. Configuration is optional: without one it scans the current directory. Put JEV_API_KEY in a .env in the directory you run from, and exported shell variables win over that file.
Exit 0 means no findings and no analysis failures, though the summary can still list skipped files. Exit 1 means at least one file has a finding at or above the threshold. Exit 2 is a configuration or analysis failure. JSON reports carry plugins, model, threshold, results and summary, with per-file failures inside results and fatal errors reported as an error object with a message rather than a result row.
Write the convention as a yes/no question, then export a plugin object from index.mjs or index.ts with an id, an instructions string or a structured object or array, and a message the agent can act on. Register the containing directory, or the file directly, under plugins in the config; the path is relative to the config file. Plugin IDs must be unique within a configuration, and each plugin can carry its own files and ignore globs.
More Jev resources like this
48GitHub stars
Agent harness for work that outlives one chat, with resumable turns, routed state machines, and observational memory across sessions and processes.
Jev Agents & MCP#community#jevbest#catalog-import
Community
dzhng/duet-agentTypeScript
dzhng/duet-agentTypeScript37GitHub stars
Assistant framework that stores specs, memory, and specialist agents as plain markdown files in your repo for Claude Code, Codex, Opencode, and similar runtimes.
Jev Agents & MCP#community#jevbest#catalog-import
36GitHub stars
Obsidian plugin providing multi-provider chat, vault read and write with function calling, visual workflow automation, and local vector search.
Jev Agents & MCP#community#jevbest#catalog-import
Community
takeshy/obsidian-llm-hubTypeScript
takeshy/obsidian-llm-hubTypeScript36GitHub stars
Pi extension that strips the full Agent Skills catalog from every request and loads only the skills a Jev ranking call judges relevant.
Jev Agents & MCP#community#jevbest#catalog-import
102GitHub stars
Behavioral code search that asks Jev to score repository fragments for a plain-language question and returns exact excerpts with paths and lines.
Jev Agents & MCP#community#jevbest#catalog-import
Community
nassim-arifette/jevgrepTypeScript
nassim-arifette/jevgrepTypeScript44GitHub stars
Pi extension that keeps a fast Executor model on implementation and brings in a separate Advisor model for consequential choices and reviews.
Jev Agents & MCP#community#jevbest#catalog-import
Community
philipbrembeck/pi-advisorTypeScript
philipbrembeck/pi-advisorTypeScriptFrom the Jev community
Posts from builders shipping with Jev right now.