# Rockmoon — Cross-Market Financial Data for AI Agents (MCP)

**English** · [中文](README.zh-CN.md) · [한국어](README.ko.md) · [日本語](README.ja.md)

Rockmoon is an **MCP server** that gives your AI agent clean, **cross-market (US / Japan / Korea)** financial data — financials, segments, ownership, valuation metrics and prices — where **every value traces back to its source filing** (SEC / EDINET / DART).

> One API key works for **both REST and MCP**. Built MCP-native, so Claude, Cursor, and any MCP client can call it directly.

- **Endpoint:** `https://mcp.rockmoon.ai/mcp` (Streamable HTTP)
- **Auth:** `X-API-Key: <your key>` header (also accepts `Authorization: Bearer <key>`)
- **Get a key:** https://rockmoon.ai
- **Official MCP registry:** `ai.rockmoon/financial-data`

## Why Rockmoon

| | Rockmoon | Typical "global" data APIs |
|---|---|---|
| **Asia depth (JP / KR)** | Deep, cleaned, filing-grade | Shallow / auto-aggregated |
| **Provenance** | Every value traces to the source filing | Rarely |
| **MCP-native** | Yes — agents call it directly | Mostly REST-only |
| **Cross-market comparability** | One canonical 80-field schema | Per-vendor mess |

Most "global" data is wide but shallow in Asia. Rockmoon is the opposite: **deep, clean, comparable Asia + US data your agent can cite.**

## Quickstart

Add to your MCP client config (Claude Desktop, Cursor, etc.):

```json
{
  "mcpServers": {
    "rockmoon": {
      "type": "http",
      "url": "https://mcp.rockmoon.ai/mcp",
      "headers": { "X-API-Key": "<YOUR_ROCKMOON_API_KEY>" }
    }
  }
}
```

Then just ask:
- *"Compare gross margins of AAPL, 7203 (Toyota) and 005930 (Samsung)."*
- *"Screen US semiconductors by ROE."*
- *"Show Toyota's revenue by segment over the last 5 years."*

## Capabilities — 18 tools

**Discovery & profile**
| Tool | What it does |
|---|---|
| `search_companies` | Resolve a ticker or name to a company across US / JP / KR |
| `get_company` | Company profile — name, exchange, industry, GICS sector/group, regulator id |

**Financials** — canonical 80-field schema, traceable to filings
| Tool | What it does |
|---|---|
| `get_financials` | Full statements bundle |
| `get_income_statement` | Income statement, by period |
| `get_balance_sheet` | Balance sheet, by period |
| `get_cash_flow` | Cash-flow statement, by period |
| `get_segments` | Revenue by product / geography / business segment |
| `get_segments_trend` | Segment revenue over time |

**Valuation & screening**
| Tool | What it does |
|---|---|
| `get_metrics` | Latest valuation + quality metrics (PE, PB, PS, EV/EBITDA, ROE, ROA, margins, market cap) |
| `get_metrics_history` | Historical TTM metrics per period |
| `screen_companies` | Cross-market screener — valuation/quality filters **plus GICS sector / industry-group** |
| `compare_metrics` | Side-by-side metrics across tickers (cross-market) |
| `get_prices` | Price history |

**Ownership & insiders**
| Tool | What it does |
|---|---|
| `get_major_shareholders` | Major shareholders |
| `get_director_holdings` | Director / officer holdings |
| `get_insider_trades` | Insider transactions |
| `get_institutional_holdings` | Institutional holdings |
| `get_large_holdings_events` | Large-holding change events |

## Provenance you can cite

Every financial value is backed by raw XBRL provenance with **ASC / IFRS / Reg S-X citations**, exposed via the REST `/v1/trace/` surface — so an agent can show *exactly* where a number came from. No black-box numbers.

## Links

- Website / get an API key — https://rockmoon.ai
- Official MCP registry — `ai.rockmoon/financial-data`