Parkour Mcp

Parkour Mcp is a Model Context Protocol (MCP) server and toolkit designed to surface high-signal, unsummarized web content for Large Language Models (LLMs). It emphasizes structured content extraction, citation awareness, and fast, targeted access to web and API data via integrations with sources like Kagi, Semantic Scholar, arXiv, GitHub, MediaWiki, and more. Built with agentic AI workflows in mind, Parkour enhances LLM toolchain performance by providing instructional frontmatter and fast content retrieval, making it highly valuable for research and agentic automations.