Overview

The ShareSeer MCP Server exposes a Model Context Protocol (MCP)–compatible toolserver that provides programmatic access to up-to-date SEC filings, company financials, and insider trading activity. It is designed to be run as a companion toolserver for LLMs or other agent frameworks that consume external facts and documents at runtime, enabling model prompts to reference authoritative market filings and transaction data without embedding large document caches inside the model.

For developers, the server is useful for building apps that require reliable, near real-time market context: regulatory monitoring, research assistants, trade signal pipelines, or compliance workflows. Because it implements MCP discovery and tool endpoints, it can be integrated quickly with agents that support the MCP discovery flow and requires only simple HTTP requests to retrieve structured filings, metadata, and time series.

Features

MCP-compatible discovery endpoint (/.well-known/mcp.json) for automated tool integration
Search and retrieval of SEC filings (10-K, 10-Q, 8-K, S-1, etc.)
Structured financials and historical metrics (income statement, balance sheet, cash flow)
Insider trading and Form 4 activity with transaction metadata
Company and CIK lookup (ticker ↔ CIK resolution)
Lightweight HTTP JSON API designed for agent/tool consumption
Configurable caching and rate-limiting to handle SEC data access safely

Installation / Configuration

Clone the repository and run with Docker, or run directly if you prefer building locally.

Clone the repo:

git clone https://github.com/shareseer/shareseer-mcp-server.git
cd shareseer-mcp-server

Build and run with Docker (recommended):

# Build the image locally
docker build -t shareseer-mcp-server .

# Run container with environment variables
docker run -d \
  -p 8080:8080 \
  -e PORT=8080 \
  -e SEC_DATA_SOURCE="edgar" \
  -e CACHE_TTL=3600 \
  -e RATE_LIMIT_REQS_PER_MIN=60 \
  shareseer-mcp-server

Minimal environment variables (examples):

PORT — port to listen on (default 8080)
SEC_DATA_SOURCE — source for filings (e.g., edgar)
CACHE_TTL — seconds to cache responses
RATE_LIMIT_REQS_PER_MIN — server-side API rate limit

If you expose the server to an LLM agent, configure TLS or run behind a reverse proxy. The server publishes an MCP discovery document at /.well-known/mcp.json for automatic tool discovery.

GitHub: https://github.com/shareseer/shareseer-mcp-server

Available Resources

Common HTTP endpoints (paths, parameters and example usage):

Endpoint	Method	Parameters	Description
/.well-known/mcp.json	GET	—	MCP discovery document for agents
/api/search/companies	GET	q (ticker or name)	Company lookup and ticker ↔ CIK resolution
/api/filings	GET	ticker, cik, form, limit, since	List filings matching filters
/api/filings/{accession}/document	GET	accession	Retrieve raw filing document or section
/api/financials	GET	ticker, period, years	Structured financials time series
/api/insiders	GET	ticker, cik, since	Insider transactions and metadata

Example: fetch latest filings for AAPL

curl "http://localhost:8080/api/filings?ticker=AAPL&limit=5"

Example: fetch insider trades for a company since 2024-01-01

curl "http://localhost:8080/api/insiders?ticker=AAPL&since=2024-01-01"

Use Cases

Augment an LLM financial assistant: Configure your agent to discover the ShareSeer MCP tool, then prompt the LLM to reference filings when answering questions like “What risk factors did Company X disclose in its latest 10-K?”
Real-time monitoring for compliance: Poll the /api/insiders endpoint to surface new Form 4 filings to compliance teams, with webhook or alerting integration for suspicious activity.
Quant research and signal generation: Combine structured financial time series from /api/financials with market data to compute fundamental ratios and backtest strategies.
Due diligence automation: Retrieve the latest S-1 or 8-K documents for a target company and feed relevant sections into a document-ranking pipeline for analyst review.

Notes & Best Practices

Respect rate limits and caching: the server may proxy or scrape external sources with their own limits (EDGAR); use CACHE_TTL and RATE_LIMIT settings to avoid overloading upstream services.
Secure the server in production: require authentication for sensitive endpoints, deploy behind HTTPS, and limit network access as appropriate.
Validate MCP discovery: confirm the MCP JSON is reachable by agent tooling at https://your-host.example/.well-known/mcp.json so automated integration can succeed.

For development, contributions, and bug reports, see the project on GitHub: https://github.com/shareseer/shareseer-mcp-server

ShareSeer MCP Server: SEC Filings & Insider Trading