Allowance MCP

Allowance MCP

Enables AI agents to request purchase approval from humans, receive scoped virtual cards, complete checkout, and report receipts for audit.

Category
访问服务器

README

Allowance MCP

<p> <a href="https://useallowance.com"> <img src="assets/allowance-logo-green-bg.jpg" alt="Allowance" width="96"> </a> </p>

Allowance is a payment security layer for AI agents. It lets humans approve scoped purchases on their phone while agents complete checkout with limited-use payment credentials instead of the human's real card.

Allowance MCP exposes that flow to MCP clients: agents can request approval for a specific purchase, receive a scoped virtual card after approval, complete checkout, and report receipts and final charged amounts back into an audit trail.

This repository is the public listing and configuration home for the hosted Allowance MCP server. The implementation repository is private.

Allowance MCP social preview

Hosted Server

  • Server URL: https://mcp.useallowance.com
  • Transport: streamable-http
  • Official registry name: com.useallowance/allowance-mcp
  • Official registry URL: https://registry.modelcontextprotocol.io/v0.1/servers/com.useallowance%2Fallowance-mcp/versions/latest
  • Company site: https://useallowance.com
  • Discovery metadata: https://useallowance.com/.well-known/mcp.json
  • Product docs: https://useallowance.com/mcp

Authentication

Allowance MCP uses OAuth. Connect your MCP client to https://mcp.useallowance.com and complete the Allowance OAuth flow when prompted.

After OAuth completes, MCP clients send the issued bearer token on tool calls. Users should not create or paste standalone API keys for normal MCP use. Do not commit tokens, card credentials, or checkout PII.

Core Flow

  1. The agent creates an allowance request for a specific purchase.
  2. The human approves, denies, or lets the request expire on their phone.
  3. If approved, the agent completes checkout with scoped payment credentials or waits for Allowance remote fulfillment.
  4. The agent reports success or failure, including receipt and charged amount details when available.

Human phone approval is the spending authorization. After approval, agents should not ask for another final confirmation unless the realized total exceeds the approved cap, checkout requires unavailable information, or the item materially differs from the approved request.

What Allowance Controls

  • Spend scope: merchant, amount cap, purchase reason, and expiration.
  • Payment exposure: agents get limited payment credentials instead of the human's real card.
  • Checkout data: approved contact, shipping, and billing fields are available only when needed for checkout.
  • Auditability: requests, approvals, card issuance, purchase attempts, receipts, and failures are recorded.
  • Revocation: humans can pause or revoke allowances instead of leaving broad payment access with an agent.

Tools

  • send_allowance_request
  • poll_allowance_request
  • wait_for_allowance_request
  • get_allowance_request
  • list_allowance_requests
  • cancel_allowance_request
  • get_contact_details
  • get_allowance_shipping_address
  • get_allowance_billing_address
  • issue_allowance_card
  • complete_purchase_auth
  • report_purchase_success
  • report_purchase_failure
  • list_allowance_purchase_attempts
  • get_allowance
  • wait_for_remote_fulfillment
  • reply_to_remote_fulfillment_message
  • list_allowances
  • pause_allowance
  • revoke_allowance
  • list_supported_merchants
  • check_merchant

Security Model

Allowance is designed so agents spend with scoped authorization instead of receiving the human's real card. Approvals, limits, receipts, and audit trails are enforced by Allowance.

If you find a security issue, email security@useallowance.com.

推荐服务器

Baidu Map

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。

官方
精选
JavaScript
Playwright MCP Server

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。

官方
精选
TypeScript
Magic Component Platform (MCP)

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。

官方
精选
本地
TypeScript
Audiense Insights MCP Server

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。

官方
精选
本地
TypeScript
VeyraX

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。

官方
精选
本地
graphlit-mcp-server

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。

官方
精选
TypeScript
Kagi MCP Server

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。

官方
精选
Python
e2b-mcp-server

e2b-mcp-server

使用 MCP 通过 e2b 运行代码。

官方
精选
Neon MCP Server

Neon MCP Server

用于与 Neon 管理 API 和数据库交互的 MCP 服务器

官方
精选
Exa MCP Server

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。

官方
精选