JevGrep

Find code by what it does. JevGrep helps coding agents find relevant code when they do not know the file name or symbol to search for.

Ask a question such as “Where is session expiry handled?” and JevGrep scans the authorized repository, asks Jev to score all eligible fragments, then returns the original source excerpts with their paths and line numbers. The calling agent can read those files in detail and continue its work with less exploratory context.

Use the CLI or connect a coding agent through the local MCP server.

Demos

CLI

MCP

What it is for

JevGrep is useful when a coding agent needs to:

• locate behaviour without knowing the exact identifier;
• understand a feature spread across implementation, configuration and tests;
• reduce the amount of repository exploration placed in the agent's main context;
• retrieve exact source excerpts instead of a generated summary.

It complements exact tools such as rg. If you already know the symbol or literal, ordinary text search is usually faster.

Requirements

• Node.js 24
• npm
• a TypeSafe AI, Vercel AI Gateway or OpenRouter API key

JevGrep searches every valid UTF-8 text file, regardless of repository language or extension.

Install

Install the public package from npm:

npm install -g @nassim-arifette/jevgrep
jevgrep --version

The unscoped package name jevgrep belongs to a different project. Use the complete scoped name above when installing. The installed command is still jevgrep.

Package: @nassim-arifette/jevgrep

To install a development checkout instead:

git clone https://github.com/nassim-arifette/jevgrep.git
cd jevgrep
npm ci
npm run build
npm link
jevgrep --version

npm link makes the jevgrep command available from any directory on the computer.

Quick start

1. Configure a provider

Configure the provider and key once for the computer:

jevgrep init --global

TypeSafe AI is proposed first. To use Vercel AI Gateway instead:

jevgrep init --global --provider vercel

To use OpenRouter:

jevgrep init --global --provider openrouter

The command stores the credential in the user's JevGrep configuration directory, not in a repository. TYPESAFE_API_KEY, AI_GATEWAY_API_KEY and OPENROUTER_API_KEY environment variables take priority over the corresponding stored value.

2. Authorize a repository

Run init once from the repository root:

cd path/to/my-project
jevgrep init

The default root is the current directory. You can also provide it explicitly:

jevgrep init --root path/to/my-project

Provider credentials are global, but repository authorization is not. Each repository must be authorized separately. Its trusted profile is stored outside the repository. Interactive init asks before enabling remote evaluation for this repository:

Allow sending eligible source excerpts from this repository to Vercel AI Gateway? [y/N]

Answer y to search immediately. Enter or n keeps remote evaluation disabled. Non-interactive initialization also leaves new profiles disabled. Optional scan caps are disabled by default; configure them if you want to limit usage. init also creates a commented .jevgrepignore in the repository when one does not already exist. Existing exclusions are preserved; .gitignore is already respected.

3. Inspect before sending code