SEC EDGAR MCP is an open-source Model Context Protocol server that connects AI models to the SEC's EDGAR filings. Built around the MCP standard, it exposes a comprehensive set of EDGAR tools and leverages the Edgartools Python library to fetch data from official SEC sources, with direct XBRL parsing for exact financial precision. Available tools include Company Tools (CIK lookup, company information, and company facts), Filing Tools (recent filings, filing content, 8-K analysis, and section extraction), Financial Tools (financial statements with direct XBRL parsing for exact precision), and Insider Trading Tools (Form 3/4/5 analysis with detailed transaction data). The server accesses data via the SEC EDGAR REST API and downloads complete filings (.txt) for content parsing, applying deterministic processing to ensure results are based solely on official data and including clickable filing references for verification. It uses flexible XBRL extraction (pattern-based) rather than hardcoded mappings to extract financial concepts. By default, it uses stdio transport, with potential integration to streamable HTTP; note that there is no authentication, so deployment should be secured in private networks.