bench get-app https://github.com/nishanthabimanyu/frappe_x402

[![Listed on Frappe Gems](https://frappegems.com/api/method/frappe_gems.seo.badge?app=nishanthabimanyu%2Ffrappe_x402)](https://frappegems.com/gems/apps/nishanthabimanyu/frappe_x402)

frappe_x402

A high-performance economic gateway for the Model Context Protocol (MCP), powered by the Frappe Framework and the x402 payment protocol.

Overview

frappe_x402 provides a robust marketplace and monetization layer for AI tools. It enables developers to register MCP-compliant services and monetize them using the x402 protocol, allowing autonomous AI agents to discover, invoke, and pay for capabilities on a per-call basis. By bridging the gap between Web2 payment gateways and Web3 settlements, it facilitates seamless "Agentic Commerce."

Core Features

• Economic Interceptor: A specialized middleware that validates user credits and triggers x402 payments before proxying requests to tool providers.
• MCP Standard Compatibility: Built specifically for the Model Context Protocol, providing standardized tool discovery and execution.
• Fiat-to-USDC Abstraction: Integrated support for Razorpay to handle fiat top-ups, which are internally tracked as credits and settled as USDC on the Base blockchain.
• Autonomous Governance: Granular control over AI agent behavior through daily spending limits and tool whitelisting.
• Unified Ledger: A persistent audit trail of all tool invocations, credit deductions, and on-chain transaction hashes.

Project Development Checkpoints

The following milestones have been achieved in the current implementation:

• Phase 1: Foundation

  • [x] Application scaffolding and workspace initialization.
  • [x] Implementation of core DocTypes: MCP Provider, MCP Tool, Workspace Credit, Credit Transaction.
  • [x] Database migration and multi-tenant schema verification.

• Phase 2: Economic Gateway

  • [x] Development of the Economic Interceptor middleware.
  • [x] Integration of x402 protocol for USDC settlement (Base Mainnet ready).
  • [x] Implementation of Razorpay fiat-to-credit bridge.
  • [x] Authenticated API access via Frappe Key/Secret system.

• Phase 3: User Interface

  • [x] Design and implementation of the Marketplace "App Store" UI.
  • [x] Deployment of the x402 Marketplace Workspace (Frappe v15 compatible).
  • [x] Real-time credit balance tracking and transaction audit logs.

• Phase 4: Operational Reality (Current)

  • [x] Integration of real-world tool providers (Web Search, English Dictionary, Crypto Prices).
  • [x] Verification of the end-to-end "Agentic Transaction" loop via demo_agent.py.
  • [x] Robust error handling and automated system vitality checks.

Technical Architecture

The system operates as a managed economic proxy:

1. Request: An AI agent sends an authenticated MCP request to the Frappe Gateway.
2. Authorization: The Gateway verifies the API credentials, user credit balance, and agent-specific policies.
3. Settlement: If authorized, the system deducts internal credits and initiates an x402 settlement to the provider's wallet.
4. Proxying: The request is forwarded to the provider's endpoint, and the response is returned to the agent.
5. Auditing: All metadata and financial logs are recorded in the Frappe database.

Prerequisites

• Frappe Framework v15
• Python 3.10+
• Node.js 18+
• Docker (recommended for development)

Installation

Install the application using the Frappe Bench CLI:

bench get-app https://github.com/nishanthabimanyu/frappe_x402.git
bench --site [your-site-name] install-app frappe_x402
bench migrate

Optional: EVM Integration

To enable on-chain USDC settlements, install the required EVM dependencies:

./env/bin/pip install "x402[evm]"

Configuration

API Authentication

Generate API keys for your AI agents through the User DocType in the Frappe Desk. Authenticate requests using the following header: Authorization: token [API_KEY]:[API_SECRET]

Payment Gateways

Configure Razorpay or Stripe credentials in the system environment variables or through the provided configuration DocTypes to enable fiat credit purchases.

Development and Testing

The repository includes a demo_agent.py script to simulate an AI agent interaction. To run the validation suite:

bench --site [your-site-name] run-tests --app frappe_x402

Contributing

We welcome contributions that align with the project's architectural standards. Please refer to CONTRIBUTING.md for our coding conventions and pull request process.

License

This project is licensed under the MIT License. See the LICENSE file for details.