Product Requirements Document

Historical note: this page is retained as an archive. For current behavior, rely on the live docs, README, and source code.

Product intent

MCP-Crawl4AI exists to connect AI systems to live web content through a focused Model Context Protocol server built on Crawl4AI and FastMCP.

Primary goals

• Deliver a reliable MCP server with clear tool, resource, and prompt contracts
• Expose high-value crawling and scraping workflows for AI assistants
• Provide outputs that are useful for LLM reasoning and downstream automation
• Keep deployment and local development straightforward for contributors

Core capabilities covered by the archived PRD

• Canonical scrape and crawl operations
• Structured extraction and conversion options
• Session-aware workflows and artifact retrieval
• Multiple transport options for clients and tools
• Performance, reliability, and safety expectations for production use

Current sources of truth

• Features
• Architecture
• API Reference
• Roadmap
• README