Jevcrawl

Jevcrawl is an MCP server and CLI that crawls websites for AI agents by using a small decision model.

Jevcrawl is a web crawler and Model Context Protocol server designed for AI agents. It uses a small decision model to pick links and passages so an LLM reads only relevant content. The system crawls websites to find specific answers without burning tokens or requiring expensive LLM-driven navigation. It works with Claude Code, Claude Desktop, Cursor, and the command line interface. Inputs include target URLs and goals, and it outputs verbatim passages with source URLs, heading paths, and character offsets.

Jevcrawl at a glance

Best for: Software engineer
Runs on: Web, CLI, API
Pricing: Jev charges $0.042 per million input tokens for jev-1.13.0
Released: October 2026
Category: AI coding tool
Website: github.com/rayenmlayeh/jevcrawl

Key features

Crawl websites for AI agents using a decision model
Return verbatim passages with sources and offsets
Support MCP clients like Claude Code and Cursor
Provide a command line interface for crawling
Cache answers locally to avoid duplicate costs
Protect against prompt injection by design

Use it for

Find specific answers deep inside documentation websites
Scrape targeted pages based on a specific goal
Map and rank URLs to decide what content to scrape

Companions

Tools that articles about Jevcrawl recommend using with it, each with its job.

Claude — Built to work with Claude.
Composio — Equips AI agents with tools, managed user auth, and sandboxes across 1,500+ apps
Make — Runs and manages Make scenarios from a conversation.
Canva — Templates for almost any design, plus AI editing in one tool
Figma — Brings design files and components into the chat.
Context7 — Up-to-date API documentation for coding questions.

Alternatives to Jevcrawl

jevgrep — Reduces coding agent costs by providing targeted, relevant repository context.
ContextMCP — Maintains fresh, accurate context for AI agents using AST-aware chunking.
G Maps Crawl — Connects your AI agents directly to live Google Maps business search data.
Firecrawl — Provides direct web data access for AI agents with open source flexibility.
jCodeMunch MCP — Cuts AI code-reading token costs by over 95 percent via tree-sitter symbol indexing.
CrawlForge MCP — Provides 31 MCP-native web scraping tools with transparent credit pricing.
claude-mem — It shares persistent, searchable agent memory across sessions and devices.
Marmot — Consolidates multiple AI and data provider interfaces into one CLI.
Lobster — Runs AI coding agents in isolated worktrees and nudges you only for human reviews.
Claude Code Router — It centralizes control of AI agent requests with intelligent routing and observability.
Codebase Memory MCP — High-speed indexing and persistent knowledge graphs for AI coding agents.
Crit — It keeps human feedback localized and precise for AI agent code iterations.

All Jevcrawl alternatives