SNACS Data MCP server

SNACS MCP Server

Connect your AI assistant to live SEC dilution forensics, market data, and news for U.S. equities updated in real-time.

https://mcp.snacs.trade - remote MCP server, HTTP transport, no local install.

What you can ask

• "Pull the full dilution chain state for TICKER - active shelfs, ATMs, warrants, converts."
• "What was this company's cash, float, and active facilities on 2024-12-31?" (point-in-time, no lookahead bias)
• "Screen for stocks under $5 with an active ATM and less than 6 months of runway."
• "Summarize today's 8-Ks and offering-related news for my watchlist."

Setup

You need an API key: snacs.trade/api -> subscribe -> account settings -> API keys (snacssklive_...).

On claude.ai (web), no key is needed: Settings -> Connectors -> Add custom connector -> URL https://mcp.snacs.trade (leave the advanced fields empty) -> Connect, then sign in with your SNACS account and approve. The server is an OAuth 2.1 protected resource (PKCE S256, DCR); clients that speak MCP OAuth discover it automatically. API keys keep working everywhere else.

The server accepts the key two ways: Authorization: Bearer snacssk... (shown below) or a plain X-API-Key: snacssk... header (useful for gateways and directories that inject a single key header).

Claude Desktop

Settings -> Connectors -> Add custom connector:

• URL: https://mcp.snacs.trade
• Header: Authorization: Bearer snacssklive_...

Claude Code

claude mcp add --transport http snacs https://mcp.snacs.trade \
  --header "Authorization: Bearer snacs_sk_live_..."

Cursor

.cursor/mcp.json:

{
  "mcpServers": {
    "snacs": {
      "url": "https://mcp.snacs.trade",
      "headers": { "Authorization": "Bearer snacs_sk_live_..." }
    }
  }
}

VS Code (Copilot)

.vscode/mcp.json:

{
  "servers": {
    "snacs": {
      "type": "http",
      "url": "https://mcp.snacs.trade",
      "headers": { "Authorization": "Bearer snacs_sk_live_..." }
    }
  }
}

Tools (14)

Pulled from the live server's tools/list (Aug 11 2026, server v3.4.2):

Shortened. The full README is on GitHub.