sec-edgar-mcp: MCP server connecting EDGAR filings to LLM workflows
sec-edgar-mcp, created by Stefanoamorelli, is an MCP server that gives AI models structured access to the U.S. SEC EDGAR system for financial research and verification. The tool enables programmatic company discovery, filing retrieval, and extraction of numeric facts so models can answer technical questions with sourceable evidence. Key capabilities include targeted extraction of filing sections, XBRL parsing, insider-transaction access, and direct URLs to filings. It targets analysts, quantitative researchers, investment teams, and developers building LLM-backed financial applications.
It turns regulatory filings into model-ready context for precise research
The tool acts as a bridge so LLMs can run research tasks without manual scraping, supporting company discovery, filing lookup by CIK or ticker, and retrieval of specific report sections. It supports tasks such as corporate performance checks, regulatory-compliance queries, and tracking insider transactions by exposing filing content in a format an assistant can ingest for focused answers.
It produces verifiable numeric outputs with direct source links
Responses include direct URLs to the original SEC filings, a measure intended to reduce hallucinations by enabling verification. The server performs XBRL extraction to pull exact numeric facts from interactive data filings, which helps produce answers that reference specific line items and filing passages rather than paraphrased summaries.
It requires MCP clients and basic developer setup but integrates with Python tools
Deployment fits developer workflows: the server is built on the edgartools Python library and runs via Docker, pip, or uv. It is compatible with MCP-capable clients such as Claude Desktop and Cursor. Configuration requires a valid User-Agent string (name and email) to comply with the SEC fair-access policy, so administrators must supply that value before queries are allowed.
It is optimized for token-efficient grounding but assumes developer resources
The design reduces token consumption by roughly 10–20x by extracting targeted sections instead of streaming whole filings into the model, which can lower context bloat in LLM prompts. That efficiency suits teams integrating citation-backed regulatory data into applications, while independent analysts without developer support may find initial setup and MCP integration demanding.
Practical choice for developer teams needing citation-backed SEC data
For teams that build LLM-backed financial tools, sec-edgar-mcp provides a practical way to ground outputs in regulatory filings and reduce context volume. Its reliance on MCP-compatible clients and a Python deployment path means it suits technical users; pairing generated answers with a quick check of the linked filing remains a prudent workflow step for high-stakes decisions.
