dino-markets-mcp
Bridges matched Kalshi and Polymarket prediction-market data to AI agents via MCP, enabling market lookup, cross-venue spread analysis, and arbitrage detection.
README
dino-markets-mcp
Stdio bridge for dino.markets — matched Kalshi + Polymarket prediction-market data for AI agents. The same event matched across both venues, cross-venue spreads, and confirmed arbitrage sized to fillable depth, for Claude, Cursor, and any MCP (Model Context Protocol) client. Market data only — not a broker; not financial advice.
Thin proxy: the bridge holds no tool definitions or schemas. Every tool is served live from the
api.dino.markets HTTP MCP endpoint, so descriptions and schemas are always up to date.
Prefer a remote connection? The server also runs directly at https://api.dino.markets/v2/mcp/
(streamable HTTP) and is listed in the official MCP registry
as markets.dino/mcp — no install needed:
claude mcp add --transport http dino-markets https://api.dino.markets/v2/mcp/
Quick start
{
"mcpServers": {
"dino-markets": {
"command": "npx",
"args": ["-y", "dino-markets-mcp"],
"env": { "DINO_API_KEY": "your-key-here" }
}
}
}
An API key is required. Get one at https://dino.markets/pricing — the Free tier includes MCP access.
Tools
| Tool | What it does |
|---|---|
find_markets |
free-text lookup by team, competition, context, or market title |
list_markets |
the matched cross-venue catalog as canonical Market objects, filterable by category |
find_arbitrage |
confirmed cross-venue Opportunity objects, filterable by sport |
get_market |
one market by id, with settlement disclosure |
get_spreads |
cross-venue price gaps on matched pairs, with verified confidence |
list_categories |
normalized categories, market types, and contextual facets |
report_bad_arb |
flag an Opportunity that looks wrong |
watch_markets |
mint a ticket for the real-time WebSocket stream and return the ws_url |
Env
| Variable | Required | Default | Description |
|---|---|---|---|
DINO_API_KEY |
Yes | — | Bearer token for API access. |
DINO_BASE_URL |
No | https://api.dino.markets |
Override API origin (testing / staging). |
DINO_TIMEOUT |
No | 10000 |
Request timeout in milliseconds. |
Development
npm install
npm run build # tsc -> dist/
Docs
Full API reference and pricing: dino.markets/docs — REST and WebSocket are also available if you'd rather skip MCP, with the OpenAPI spec at api.dino.markets/openapi.json.
MIT © Nusantara Ventures LLC
推荐服务器
Baidu Map
百度地图核心API现已全面兼容MCP协议,是国内首家兼容MCP协议的地图服务商。
Playwright MCP Server
一个模型上下文协议服务器,它使大型语言模型能够通过结构化的可访问性快照与网页进行交互,而无需视觉模型或屏幕截图。
Magic Component Platform (MCP)
一个由人工智能驱动的工具,可以从自然语言描述生成现代化的用户界面组件,并与流行的集成开发环境(IDE)集成,从而简化用户界面开发流程。
Audiense Insights MCP Server
通过模型上下文协议启用与 Audiense Insights 账户的交互,从而促进营销洞察和受众数据的提取和分析,包括人口统计信息、行为和影响者互动。
VeyraX
一个单一的 MCP 工具,连接你所有喜爱的工具:Gmail、日历以及其他 40 多个工具。
graphlit-mcp-server
模型上下文协议 (MCP) 服务器实现了 MCP 客户端与 Graphlit 服务之间的集成。 除了网络爬取之外,还可以将任何内容(从 Slack 到 Gmail 再到播客订阅源)导入到 Graphlit 项目中,然后从 MCP 客户端检索相关内容。
Kagi MCP Server
一个 MCP 服务器,集成了 Kagi 搜索功能和 Claude AI,使 Claude 能够在回答需要最新信息的问题时执行实时网络搜索。
e2b-mcp-server
使用 MCP 通过 e2b 运行代码。
Neon MCP Server
用于与 Neon 管理 API 和数据库交互的 MCP 服务器
Exa MCP Server
模型上下文协议(MCP)服务器允许像 Claude 这样的 AI 助手使用 Exa AI 搜索 API 进行网络搜索。这种设置允许 AI 模型以安全和受控的方式获取实时的网络信息。