dartpoint

The DartPoint MCP server exposes Korean corporate data from DART (Financial Supervisory Service) via the MCP protocol, enabling LLMs and other clients to perform tasks such as financial analysis, competitor profiling, and sentiment monitoring. It supports standard MCP interactions, so you can route queries through supported clients (Claude, Cursor, and others) to retrieve structured corporate information, financial summaries, market sentiment, and product/service overviews. Examples include analyzing recent financial performance, identifying main competitors and their products, and obtaining current market sentiment from news sources. To use it, configure your MCP client to point at the server URL and provide your API key as a header (DARTPOINT_API_KEY) or environment variable as required by your client setup. The repository examples show how to connect via SSE endpoints and stdio via SuperGateway, returning structured responses suitable for downstream analysis and report generation.

Prerequisites:

• Node.js (recommended: LTS, v14+ or newer) and npm installed on your system
• Access to the internet to fetch the MCP remote package via npx

Installation steps:

1. Ensure Node.js and npm are installed
   • Inspect versions: node -v and npm -v
2. Obtain your DartPoint API key from https://dartpoint.ai
3. Create an MCP client config file (optional but recommended). Example content:

"{ "mcpServers": { "dartpoint": { "command": "npx",\n "args": [\n "mcp-remote",\n "https://dartpoint.ai/mcp-sse/mcp",\n "--header",\n "DARTPOINT_API_KEY:<API Key from https://dartpoint.ai>"\n ],\n "env": {\n "DARTPOINT_API_KEY": "Your API key from https://dartpoint.ai\"\n }\n } } }"

4. Run via an MCP client that supports the npx-based configuration (as shown in the README examples). For a quick start, you can directly run the npx command in your shell, replacing <API Key> with your actual key: npx mcp-remote https://dartpoint.ai/mcp-sse/mcp --header DARTPOINT_API_KEY:<API Key from https://dartpoint.ai>

5. Alternatively, use the provided examples in the README to integrate with Claude Desktop Standard (SSE/stdio) or Cursor-AI (SSE). Ensure your environment variables are set if your client requires them.

Tips and common considerations:

• Always keep your DARTPOINT_API_KEY secure; avoid embedding it in public repos. Prefer environment variables in your runtime environment where possible.
• The MCP URL is streamed (SSE) at https://dartpoint.ai/mcp-sse/mcp. Use the SSE variant if your client supports it for real-time responses.
• If your MCP client expects a different header format, adjust the --header value accordingly. The examples assume DARTPOINT_API_KEY is the required header.
• This server relies on the DartPoint API; API quota or key issues may affect response availability. Monitor API usage via your DartPoint account.
• If you switch clients (Claude, Cursor, SuperGateway, etc.), make sure to adapt the config structure to match the client’s expected MCP manifest (e.g., url + headers vs. full command/args).
• For debugging, you can test the endpoint with simple SSE clients or curl-based tooling that supports streaming JSON.