secapi-mcp
Enables AI clients to search and analyze SEC filings, financial statements, insider trades, and institutional holdings through natural language tools.
README
secapi-mcp
Official Model Context Protocol server for secapi.dev.
Expose SEC filings, financial statements, insider trades, and institutional holdings as AI-native tools for Claude, Cursor, ChatGPT, Windsurf, VS Code, and any MCP-compatible client.
AI Client → MCP Tools → secapi.dev REST API → JSON Response → LLM
This server never talks to databases directly — it is a typed, validated interface over the secapi-node SDK.
Features
- 9 high-level AI tools designed for reasoning, not REST endpoint mirroring
- SDK-first architecture — all API calls go through
secapi-node - Zod validation on every tool input
- AI-friendly errors with categories, suggestions, and retry hints
- Structured JSON responses with metadata and pagination
- Extensible tool registry — add new tools in minutes
- Production logging — structured, secrets-safe
- Full test suite with Vitest
- CI/CD with GitHub Actions and Changesets
Installation
npm install -g secapi-mcp
# or
pnpm add -g secapi-mcp
# or run via npx
npx secapi-mcp
Requires Node.js 20+.
Quick Start
-
Get an API key from secapi.dev
-
Set your environment variable:
export SECAPI_API_KEY="your-api-key"
- Run the server (stdio transport):
secapi-mcp
MCP Client Configuration
Cursor
Add to .cursor/mcp.json (or Cursor Settings → MCP):
{
"mcpServers": {
"secapi": {
"command": "npx",
"args": ["-y", "secapi-mcp"],
"env": {
"SECAPI_API_KEY": "your-api-key"
}
}
}
}
Claude Desktop
Add to ~/Library/Application Support/Claude/claude_desktop_config.json (macOS):
{
"mcpServers": {
"secapi": {
"command": "npx",
"args": ["-y", "secapi-mcp"],
"env": {
"SECAPI_API_KEY": "your-api-key"
}
}
}
}
VS Code (GitHub Copilot)
Add to your VS Code MCP settings:
{
"mcp": {
"servers": {
"secapi": {
"command": "npx",
"args": ["-y", "secapi-mcp"],
"env": {
"SECAPI_API_KEY": "your-api-key"
}
}
}
}
}
Windsurf
Add to Windsurf MCP configuration:
{
"mcpServers": {
"secapi": {
"command": "npx",
"args": ["-y", "secapi-mcp"],
"env": {
"SECAPI_API_KEY": "your-api-key"
}
}
}
}
Available Tools
| Tool | Description |
|---|---|
search_filings |
Search SEC filings by ticker/CIK, form type, and date range |
analyze_filing |
Get filing metadata and document list for deeper analysis |
search_companies |
Search companies by name, ticker, or CIK |
company_overview |
Company profile plus recent filings — ideal research starting point |
company_financials |
Income statement, balance sheet, or cash flow |
company_ratios |
Financial ratio time series (ROE, margins, liquidity, etc.) |
insider_trades |
Insider buying/selling activity from Form 4 filings |
search_institutions |
Find institutional investors (13F filers) |
institution_holdings |
13F portfolio holdings for a given quarter |
Environment Variables
| Variable | Required | Default | Description |
|---|---|---|---|
SECAPI_API_KEY |
Yes* | — | API key from secapi.dev |
SEC_API_KEY |
Yes* | — | Legacy alias for SECAPI_API_KEY |
SECAPI_BASE_URL |
No | https://api.secapi.dev |
API base URL |
SECAPI_TIMEOUT_MS |
No | 30000 |
Request timeout in milliseconds |
SECAPI_MAX_RETRIES |
No | 2 |
Retry count for transient failures |
SECAPI_DEBUG |
No | false |
Enable debug logging to stderr |
SECAPI_USER_AGENT |
No | secapi-mcp/0.1.0 |
User-Agent header |
* One of SECAPI_API_KEY or SEC_API_KEY is required.
Architecture
src/
├── index.ts # stdio entry point
├── server/ # MCP server factory
├── client/ # secapi-node wrapper
├── config/ # Zod-validated configuration
├── errors/ # AI-friendly error categorization
├── logging/ # Structured, secrets-safe logging
├── schemas/ # Shared Zod schemas
├── utils/ # Response formatting
└── tools/
├── registry.ts # Tool registration system
├── types.ts # Tool definition interfaces
├── filings/ # Filing tools
├── companies/ # Company tools
├── financials/ # Financial statement tools
├── insiders/ # Insider trading tools
└── institutions/ # 13F institutional tools
Adding a New Tool
- Create a tool definition in the appropriate
src/tools/<domain>/index.ts:
{
name: "my_new_tool",
description: "Clear description for the LLM...",
inputSchema: z.object({ ticker: tickerSchema }),
handler: async (input, { client }) => {
const data = await client.filings.search({ ticker: input.ticker });
return createToolResponse("my_new_tool", data.data, {
pagination: data.pagination,
});
},
}
- Export the module from
src/tools/index.tsif it's a new domain.
That's it — the registry handles validation, error formatting, and MCP registration.
Development
git clone https://github.com/secapi-dev/secapi-mcp.git
cd secapi-mcp
pnpm install
pnpm test
pnpm build
Run locally:
SECAPI_API_KEY=your-key pnpm start
Scripts
| Script | Description |
|---|---|
pnpm build |
Build to dist/ |
pnpm test |
Run Vitest test suite |
pnpm typecheck |
TypeScript type checking |
pnpm lint |
ESLint |
pnpm format |
Prettier format |
Troubleshooting
"API key is required" — Set SECAPI_API_KEY in your MCP client config env block.
"Invalid or missing API key" — Verify your key at secapi.dev. Keys are never logged.
"API rate limit exceeded" — Wait and retry, or upgrade your plan.
Server not appearing in Cursor — Restart Cursor after editing MCP config. Check stderr logs with SECAPI_DEBUG=true.
FAQ
Why not mirror REST endpoints 1:1?
AI agents work better with semantic, high-level tools. company_overview is more useful than GET /entities/:id + GET /entities/:id/filings.
Can I self-host?
Yes. Set SECAPI_BASE_URL if using a custom API deployment.
How do I add more tools? See Adding a New Tool. The framework is designed for fast extension.
License
MIT — see LICENSE.
Links
- secapi.dev — API platform
- secapi-node — TypeScript SDK
- Model Context Protocol — MCP specification
Recommended Servers
playwright-mcp
A Model Context Protocol server that enables LLMs to interact with web pages through structured accessibility snapshots without requiring vision models or screenshots.
Audiense Insights MCP Server
Enables interaction with Audiense Insights accounts via the Model Context Protocol, facilitating the extraction and analysis of marketing insights and audience data including demographics, behavior, and influencer engagement.
Magic Component Platform (MCP)
An AI-powered tool that generates modern UI components from natural language descriptions, integrating with popular IDEs to streamline UI development workflow.
VeyraX MCP
Single MCP tool to connect all your favorite tools: Gmail, Calendar and 40 more.
graphlit-mcp-server
The Model Context Protocol (MCP) Server enables integration between MCP clients and the Graphlit service. Ingest anything from Slack to Gmail to podcast feeds, in addition to web crawling, into a Graphlit project - and then retrieve relevant contents from the MCP client.
Kagi MCP Server
An MCP server that integrates Kagi search capabilities with Claude AI, enabling Claude to perform real-time web searches when answering questions that require up-to-date information.
Neon Database
MCP server for interacting with Neon Management API and databases
Exa Search
A Model Context Protocol (MCP) server lets AI assistants like Claude use the Exa AI Search API for web searches. This setup allows AI models to get real-time web information in a safe and controlled way.
Qdrant Server
This repository is an example of how to create a MCP server for Qdrant, a vector search engine.
E2B
Using MCP to run code via e2b.