DartLab is a powerful financial data MCP server designed for Korean DART and US EDGAR disclosure data. It offers rich data interfaces and AI analysis capabilities, suitable for deep financial research. Installation requires pre-installing the Python package and configuring MCP clients. It currently supports Claude Desktop, Claude Code, Codex, and Cursor. Note that it requires significant local disk space and network resources, and provides limited depth for US stocks. Overall, it is an extremely valuable tool for researchers focused on the Korean market.
Strongest · Maintenance 14/20 Weakest · Reliability 8/20
Static review · not runListed 2026-08-07
Read the FMRS scoring method →Runtime：Python
PYTHONUNBUFFERED
Keeps Python output unbuffered so MCP stdio messages arrive in real time; set in the client config, nothing to obtain.
PYTHONUTF8
Enables Python UTF-8 mode for correct handling of Korean filing text; set in the client config, nothing to obtain.
Install the Python package: pip install dartlab or uv add dartlab. Then, for Claude Desktop, add an MCP server config in your settings file using command dartlab mcp (pre-install with uv tool install dartlab). For Claude Code, run claude mcp add dartlab -- dartlab mcp. For Codex CLI, run codex mcp add dartlab -- dartlab mcp. You can also connect via remote SSE: https://eddmpython-dartlab.hf.space/mcp/sse (note: Claude Desktop does not support remote endpoints).
{
"mcpServers": {
"dartlab": {
"command": "dartlab",
"args": ["mcp"],
"env": { "PYTHONUNBUFFERED": "1", "PYTHONUTF8": "1" }
}
}
}
Shown for Claude Desktop. Other clients may use a different file or key (VS Code uses "servers") — the configurator below converts it.
{
"servers": {
"dartlab": {
"command": "dartlab",
"args": [
"mcp"
],
"env": {
"PYTHONUNBUFFERED": "1",
"PYTHONUTF8": "1"
}
}
}
}
Goes in your project's .vscode/mcp.json (VS Code uses a "servers" key).
claude mcp add dartlab -e PYTHONUNBUFFERED=1 -e PYTHONUTF8=1 -- dartlab mcp
Run it in a terminal; replace any <…> placeholders with your own values first.
After installing (`uv tool install dartlab` or `pipx install dartlab`), add the MCP config (`dartlab mcp` entry point). Once connected, tools like ask, EngineCall, and RunPython should appear in the client's tool list; asking about Samsung Electronics' income statement with ref-backed numbers confirms it works.
Once connected, you can ask your AI assistant things like:
No matching tools
Listed from the project's documentation, not tested by this site.
DartLab is an open-source research system that turns Korean DART and US SEC EDGAR disclosures into comparable corporate data. It provides financial statements, business report full text, filing lists, ratios, credit risk, industry maps, and macro context through a unified Company interface. The core is horizontalizing each company's differing account names and disclosure tables into topic × period and account × period grids, enabling comparisons across years, across companies (e.g., Samsung vs Apple), and across the entire market. The system offers a Python library, CLI, and AI assistant (via MCP) with five analysis lenses (Analysis, Credit, Industry, Quant, Macro), plus scan, screener, story, and simulation features. Data is prebuilt on HuggingFace, and no API keys are required to get started.
Upstash's official server providing up-to-date third-party library docs for AI coding assistants
AI-powered NPM package analysis MCP server
An independent control layer that gives AI coding agents budgets, brakes, and receipts
GitHub's official MCP server for managing repos, issues, PRs, and workflows via natural language
Source revision 0a570038b13f Data synced 2026-10-12 Read the FMRS scoring method