Skip to main content

Documentation Index

Fetch the complete documentation index at: https://docs.logokit.com/llms.txt

Use this file to discover all available pages before exploring further.

The LogoKit MCP server is available on Startup, Scale, and Enterprise plans. It uses the same secret API key as the Brand Data API — upgrade here if you are on the free plan.

​
What it is

LogoKit MCP is a hosted Model Context Protocol server that exposes LogoKit’s brand database to AI assistants. Your agent can look up any company by domain or by financial identifier (ticker, ISIN, CUSIP, SEDOL, WKN) and receive structured Brand Data — logo URL, brand colors, social profiles, headquarters, founding year, employee range, and GICS classification.
  • Endpoint: https://mcp.logokit.com/mcp
  • Transport: Streamable HTTP (JSON-RPC 2.0)
  • Protocol version: 2024-11-05
  • Authentication: Bearer secret key (sk_pa_… or sk_en_…)

​
Authentication

Pass your secret API key in the Authorization header: 
Authorization: Bearer sk_pa_YOUR_KEY
Only secret keys are accepted — publishable keys (pk_*) are rejected because MCP traffic is run by your agent, not an end-user’s browser. Find your secret key in the API Tokens dashboard.

​
Tools

The MCP server exposes two tools. Both return the same Brand Data record.

​
get_brand_by_domain

Look up a brand by its canonical website domain. Subdomains and www. prefixes are normalized automatically.
ArgumentTypeRequiredDescription
domainstringyesWebsite domain — e.g. stripe.com, apple.com
Example call
{
  "name": "get_brand_by_domain",
  "arguments": { "domain": "stripe.com" }
}

​
get_brand_by_ticker

Look up a brand by a financial security identifier. Accepts stock tickers, ISIN, CUSIP, SEDOL, and WKN — case-insensitive.
ArgumentTypeRequiredDescription
symbolstringyesTicker (AAPL), ISIN (US0378331005), CUSIP (037833100), SEDOL (2046251), or WKN (865985)
Example call
{
  "name": "get_brand_by_ticker",
  "arguments": { "symbol": "NVDA" }
}

​
Response shape

A successful tool call returns the Brand Data record as both pretty-printed text content and structured content: 
{
  "name": "Apple",
  "domain": "apple.com",
  "long_name": "Apple Inc.",
  "website": "https://www.apple.com",
  "logo": "https://img.logokit.com/apple.com",
  "colors": ["#000000", "#f5f5f7", "#ffffff"],
  "founded_year": 1977,
  "number_of_employees": "150,000-200,000",
  "social_media": {
    "x": "https://x.com/apple",
    "linkedin": "https://www.linkedin.com/company/apple"
  },
  "address": {
    "address1": "One Apple Park Way",
    "city": "Cupertino",
    "state": "CA",
    "zip": "95014",
    "country": "US"
  },
  "gics_sector": "Information Technology",
  "gics_industry": "Technology Hardware, Storage & Peripherals"
}
Empty and missing fields are omitted from the response. Absence does not imply the brand lacks the attribute — it means LogoKit doesn’t have the value populated yet. Your agent should present the returned data as-is and avoid speculating about missing fields.

​
Errors

Errors follow the JSON-RPC 2.0 spec. Common cases:
CodeMeaning
-32001Unauthorized — missing/invalid key, banned, inactive, or quota exceeded
-32601Method not found
-32602Invalid arguments (unknown tool, missing required arg)
-32700Parse error — invalid JSON
-32603Internal server error
When a brand cannot be found, the tool call returns an isError: true result with a descriptive message rather than a JSON-RPC error — so the agent can recover gracefully and ask the user for a different identifier.

​
Usage and quotas

MCP requests count against your Brand Data API quota. You can monitor your usage in the developer dashboard. Plan limits are 25K and 100K requests/day for Startup and Scale respectively. Token status (active, usage_exceeded, banned) is enforced on every request — see Rate Limits for details.

​
Next step

Install in your client

Copy-pasteable configs for Claude Desktop, Claude Code, Cursor, VS Code, Codex, Windsurf, ChatGPT, and any other MCP-compatible client.