Skip to main content
Glama
Korea Stock Insight MCP

Korea Stock Insight MCP

Korea Stock Insight MCP

mcp-name: io.github.yjhann33-design/korea-stock-insight-mcp

PyPI License: MIT

English-first MCP — built for global LLM analysts, not Korean fluency speakers.

Drop-in for Claude Desktop, Cursor, Cline, and any MCP-compatible client. Ask in English, get Korean filings, financials, and KOSPI/KOSDAQ prices in one tool call chain — DART corp_code lookup, disclosures, XBRL statements, daily OHLCV.

What you get

Tool

What it does

get_today_date

Today in KST/UTC (YYYYMMDD) — call before any date-range query

get_corp_code

Resolve Korean company name or 6-digit ticker → DART corp_code

get_disclosure_list

Recent DART filings (date, title, type, receipt no.)

get_disclosure

Filing metadata + canonical viewer URL

get_financial_statement

Full XBRL financial statements (consolidated/separate, quarterly/annual)

get_stock_quote

Daily OHLCV from KOSPI/KOSDAQ (via yfinance, no key required)

DART (전자공시시스템) is South Korea's electronic disclosure system — the SEC EDGAR equivalent, operated by the Financial Supervisory Service (FSS).

Related MCP server: opendart-mcp-server

Why English-first matters

Korean public filings are written in Korean, ticker codes are 6-digit numerics with no global mnemonic, and the disclosure system (DART) has no English equivalent of EDGAR's full-text search. For a non-Korean analyst, that's three friction layers before any actual research.

This server collapses them. Tool descriptions, parameter names, and examples are all written for an LLM that doesn't read Korean — company resolution accepts English names ("Samsung Electronics", "Naver"), responses include English context where useful, and the workflow is one Claude/Cursor turn instead of three browser tabs.

  • English-first tool descriptions — names, params, examples all in English

  • DART + KRX in one place — disclosures, financials, daily prices through a single MCP

  • One-line install, no account — uvx korea-stock-insight-mcp and you're done (only a free DART key needed)

  • Local stdio, MIT, no telemetry — your queries don't leave your machine

Quick start (Claude Desktop)

1. Get a free DART API key

Register at opendart.fss.or.kr and grab your API key from the user page. Instant, free, no credit card.

2. Add to your Claude Desktop config

Edit claude_desktop_config.json (location: macOS ~/Library/Application Support/Claude/, Windows %APPDATA%\Claude\):

{
  "mcpServers": {
    "korea-stock-insight": {
      "command": "uvx",
      "args": ["korea-stock-insight-mcp"],
      "env": {
        "DART_API_KEY": "YOUR_DART_KEY_HERE"
      }
    }
  }
}

If you don't have uvx, install uv first (one-liner on macOS/Linux/Windows), or use pipx:

{
  "mcpServers": {
    "korea-stock-insight": {
      "command": "pipx",
      "args": ["run", "korea-stock-insight-mcp"],
      "env": {"DART_API_KEY": "YOUR_DART_KEY_HERE"}
    }
  }
}

3. Restart Claude Desktop and try

"What were Samsung Electronics' most recent disclosures? Pull their last quarter's revenue and operating profit."

Claude will chain get_corp_code("Samsung Electronics") → get_disclosure_list("005930") → get_financial_statement(corp_code, 2025, "FY") automatically.

Sample queries (30 seconds, copy-paste into Claude)

Three workflows global analysts actually run on Korean equities. Each one chains 2–4 tool calls automatically.

1. Memory-cycle peer comparison

"Compare Samsung Electronics and SK Hynix on revenue, operating margin, and capex for the most recent annual filing. Highlight where they diverge."

Resolves both names to corp_codes → pulls latest XBRL → returns a side-by-side. Useful for tracking the HBM/DRAM cycle.

2. Disclosure radar before a catalyst

"Show me Naver's last 30 days of DART filings. Flag anything about earnings, M&A, share buybacks, or executive changes."

get_disclosure_list with date range, then Claude triages titles. No Korean reading required — the tool description tells Claude what filing types to look for.

3. Price + filing context in one turn

"What did Kakao file most recently, and how has the stock moved on KOSPI over the past 90 days?"

Combines get_disclosure_list + get_stock_quote so you get the news and the price reaction in the same answer.

Other examples

  • "Pull LG Energy Solution's last annual report and summarize the management discussion."

  • "What's the KODEX 200 ETF's 6-month trend?" (works with ETFs too)

  • "For Hyundai Motor, list every disclosure tagged as a major decision (주요사항보고서) in 2025."

Roadmap

  • v0.2 — Full disclosure body extraction (zipped XBRL → plain text)

  • v0.3 — Multi-company batch queries, sector aggregations

  • Future — Optional hosted tier with English-summarized disclosures and webhook push (no ETA; track progress in GitHub Issues)

Feedback & contributions

This is solo-maintained, and stdio mode means I can't see who's using it. The only signals I get are PyPI downloads, GitHub stars, and what you tell me.

If this helps you, please let me know:

  • 🐛 Open an issue — bugs, missing tools, broken queries

  • 💬 Discussions — "anyone analyzed [ticker]?", "is this the right tool for X?", general questions

  • ⭐ Star the repo — easiest way to signal demand and shape the roadmap

If you actually analyze Korean stocks with Claude/Cursor, I'd genuinely love to hear what queries you run and what's missing.

License

MIT. See LICENSE.

Acknowledgments

Available Tools

6 tools
get_corp_codeA

Resolve Korean company name or stock code to DART corp_code (8-digit).

DART corp_code is the unique identifier needed for financial-statement and detailed-disclosure lookups. Accepts either a 6-digit stock code (exact match) or a Korean company name (partial, case-insensitive).

For non-Korean speakers: search the company's English name first via web search, then pass the official Korean name to this tool.

Inputs: query (required).

get_disclosureB

Fetch metadata + viewer URL for a specific DART filing.

v0.1 returns the canonical dart.fss.or.kr viewer URL. v0.2 will add in-process body extraction (DART filings are zipped XBRL).

Inputs: rcept_no (required).

get_disclosure_listA

List recent DART disclosures for a Korean-listed company.

Returns filing metadata (date, title, type, receipt number) for the last days calendar days. Use rcept_no with get_disclosure to open a specific filing.

Inputs: stock_code (required), days.

get_financial_statementA

Full XBRL financial statements for a Korean company.

Returns every account line from the periodic report. Use get_corp_code first to resolve a name/ticker to the 8-digit corp_code.

Args: corp_code: 8-digit DART code. year: Business year (e.g. 2025). quarter: "Q1" / "H1" / "Q3" / "FY" (default "FY" = annual report). fs_div: "CFS" (consolidated, default) or "OFS" (separate).

Inputs: corp_code (required), year (required), quarter, fs_div.

get_stock_quoteA

Daily OHLCV bars for a Korean-listed stock (KOSPI/KOSDAQ).

Auto-detects the market by trying .KS then .KQ. Prices in KRW. v0.1 uses yfinance (no API key required); v0.2 may add official KRX OpenAPI for intraday and accurate market cap.

Inputs: stock_code (required), days.

get_today_dateA

Return today's date in KST and UTC (YYYYMMDD format).

Use before any date-range query — Korean market APIs require YYYYMMDD and KST is the trading-day timezone.

Inputs: no named parameters.

Tool Schema Changelog

Recent tool additions, removals, and schema changes observed during successful MCP inspections.

Loading tool schema changelog…

    TDQS

    A4.1/5.0

    Scored across 6 tools

    Disambiguation5/5

    Each tool has a clearly distinct purpose: company code resolution, filing metadata retrieval, filing listing, financial statements, stock quotes, and date utility. No overlap or ambiguity.

    Naming Consistency5/5

    All tool names follow a consistent 'verb_noun' pattern in snake_case (e.g., get_corp_code, get_disclosure). This makes the tool set predictable and easy to navigate.

    Tool Count5/5

    With 6 tools, the server is well-scoped for its purpose. Each tool addresses a core aspect of Korean stock data access without unnecessary bloat or missing essentials.

    Completeness4/5

    The tool set covers key workflows: company lookup, disclosure browsing and retrieval, financial statements, stock quotes, and a date utility. Minor gaps like missing a dedicated search tool are mitigated by existing partial-match capabilities.

    Maintenance

    ActivityInactive
    ResponsivenessNo recent community issues

    Related MCP Connectors

    Related MCP Servers

    • F
      license
      Not graded
      quality
      D
      maintenance
      Enables AI agents to query Korean listed companies' financial statements, public disclosures, executive information, and shareholder structures in real-time using the DART API.
      2
      -
    • A
      license
      Not graded
      quality
      D
      maintenance
      Enables AI agents to access Korean corporate disclosure data from DART, allowing natural language queries about companies, financial statements, and disclosures.
      54 npm
      2
      MIT
    • A
      license
      Not graded
      quality
      D
      maintenance
      Enables AI-powered analysis of Korean stock market data and corporate disclosures using official DART and KRX APIs.
      ISC
    • A
      license
      Not graded
      quality
      D
      maintenance
      Provides AI assistants with real-time access to Korean listed companies' disclosures, financial statements, and corporate information via the DART API.
      4
      MIT