agentbank-merchant-mcp
Merchant-side MCP server for agentbank that lets merchants read their own orders and live Curless wallet balance from an MCP client such as Claude Desktop.
README
@curless/agentbank-merchant-mcp
Merchant-side MCP server for agentbank (the "Stripe for AI agents"). Lets a merchant read their own orders and live Curless wallet balance (real funds settled into the merchant's Curless account) from an MCP client such as Claude Desktop.
Authenticated by your Curless API key via env — no OAuth. Renders an MCP Apps card (orders / balance widget) in hosts that support it, with a markdown table fallback everywhere else. Kept separate from the buyer-side MCP so the two personas don't share a tool surface.
Published to npm as
@curless/agentbank-merchant-mcp.
Tools
list_orders— this merchant's orders (what agents have paid), newest first.get_balance— the live Curless wallet balance per currency (available / frozen).
Install (Claude Desktop)
Desktop extension (.mcpb): install agentbank-merchant.mcpb and enter your
Curless API key + merchant id when prompted.
Or via claude_desktop_config.json:
{
"mcpServers": {
"agentbank-merchant": {
"command": "/usr/local/bin/npx",
"args": ["-y", "@curless/agentbank-merchant-mcp"],
"env": {
"AGENTBANK_MERCHANT_TOKEN": "<your-curless-api-key>",
"AGENTBANK_MERCHANT_ID": "888888"
}
}
}
}
| env | what |
|---|---|
AGENTBANK_MERCHANT_TOKEN |
your Curless API key (issued by Curless at onboarding) |
AGENTBANK_MERCHANT_ID |
your Curless merchant id, e.g. 888888 |
The server talks to https://mcp.curless.ai by default — no need to configure it.
(Local dev only: set AGENTBANK_API_URL=http://localhost:3000 to point at a local API.)
Build the desktop extension (.mcpb)
npm install
npm run build:mcpb # esbuild → self-contained server bundle, then `mcpb pack`
This writes mcpb/server/index.js (a single self-contained ESM bundle) and packs
agentbank-merchant.mcpb. Both are git-ignored build artifacts.
License
MIT © robin
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。