Back to feed
Dev.to
Dev.to
7/19/2026
tearsheet: a local, keyless MCP server for token-efficient web scraping in Claude Code

tearsheet: a local, keyless MCP server for token-efficient web scraping in Claude Code

Original: I built a local, keyless Firecrawl for Claude Code — here's why published: false

Short summary

tearsheet is an MIT-licensed, local-first MCP server that gives Claude Code Firecrawl-style web scraping without API keys, quotas, or data leaving your machine. It exposes five tools (scrape, search, map, crawl, extract) optimized for token efficiency by returning only main-content markdown and spilling large pages to disk. The tool is deliberately noisy about extraction failures, preferring loud omission over silent fabrication.

  • •Local, keyless MCP server replacing Firecrawl for Claude Code workflows
  • •Five tools: scrape, search, map, crawl, extract — all token-optimized
  • •Failure mode is omission not fabrication; warnings surface unreliable extractions

Generated with AI, which can make mistakes.

#open-source#ai-tools#ai-agents
Read full article at Dev.to

Is this a good recommendation for you?

Comments

Failed to load comments. Please try again.

Explore more

Scrapling MCP: Giving Claude Code a Real Web Scraper
Dev.to
Harvest: An open-source self-healing web scraper with semantic caching and MCP support
Dev.to
I Found Firecrawl Too Expensive for My AI Agent's Knowledge Base, So I Built My Own
Dev.to
The original title is: "I built an open-source MCP server so AI agents can read the web - even sites that block bots"
Dev.to