Phone Portal ↗ Preorder
⚡ testnet-stage developer demo

Agents can think.
They still cannot safely pay.

MAGNE Agent Pay is a developer-stage payment workflow demonstration for AI agents, combining M Hash L2 testnet contract deployment with verifiable AI Task Receipt records.
Currently demonstrating testnet contract deployment and AI Task Receipt recording on M Hash L2 (chainId: 20250827). Full HTTP 402 payment flow, browser wallet authorization, and Facilitator verify/settle are in active development.

View Terminal Demo View GitHub Developer Docs
20250827
Chain ID
V0.2
Contracts & Receipt Demo ✓
x402
Compatibility Under Validation
The Problem

AI agents hit a payment wall.

Current infrastructure lets agents think but not transact reliably.

401 No Secure Authorization

Agents trigger. Users can't verify.

Agents may trigger tasks, but payments still require a trusted human to authorize each step. There's no cryptographically verifiable authorization path.

0 No Agent Payment Rail

Services exist. Agents can't pay.

Agent payment rails are emerging. Secure mobile authorization and verifiable task receipts remain open implementation challenges.

0 No Verifiable Receipt

Tasks run. Nothing is provable.

AI tasks need auditable records: who requested, who paid, what settled, and what was returned. Today's workflows leave no traceable record.

The Solution

Agent Pay Workflow

From task request to verifiable receipt — a developer workflow for AI agent payment on M Hash L2.

⚠ The workflow below shows MAGNE Agent Pay's target design. Currently verified: M Hash L2 testnet contract deployment and AI Task Receipt recording. HTTP 402 payment challenge, browser wallet authorization, and Facilitator verify/settle are in development.

1
AI Agent Task
Create task via API
→
2
402 Response
Payment required metadata
→
3
Wallet Auth
MetaMask or MAGNE Phone
→
4
mMHA Transfer
ERC20 on M Hash L2
→
5
Facilitator
Strict tx verification
→
6
AI Task Receipt
On-chain, immutable
Runtime Deployed

Contract Deployment & Receipt Demo

Contract deployment and receipt recording on M Hash L2 Testnet. AI Task Receipt recording reference disclosed.

magne@agent-pay-demo M HASH L2 TESTNET DEMONSTRATION
$ magne agent task create "analyze wallet risk"
✓ task created — task_demo_20260522

$ magne api call wallet-risk --wallet 0x742d35Cc6634C0532925a3b844Bc9e7595f2E26E
✕ 402 Payment Required
amount: 0.01 mMHA
recipient: 0xf39Fd6e51aad88F6F4ce6aB8827279cffB922f66
chain: mhash-l2-testnet (20250827)
rpc: https://l2-rpc.testnet.magicalhash.com

$ magne pay settle --tx 0x0000000000000000000000000000000000000000000000000000000000000000 --network mhash-l2
✓ payment verified
status: verified
block: 8603
token: mMHA (ERC20)

$ magne receipt create --task task_demo_20260522
✓ AI Task Receipt created
receiptId: 0x07bc0ca32a037ab755b4516641fba45fe93ac33ed8864bdd9928d7ac152390fd
explorer: https://l2-explorer.testnet.magicalhash.com/tx/0x07bc0ca32a037ab755b4516641fba45fe93ac33ed8864bdd9928d7ac152390fd

$ magne receipt inspect 0x07bc0ca32a037ab755b4516641fba45fe93ac33ed8864bdd9928d7ac152390fd
agent: MAGNE-Agent-001
service: wallet-risk-analysis
amount: 0.01 mMHA
status: settled
network: M Hash L2 (chainId: 20250827)
$ █
On-Chain Verification

AI Task Receipt Layer

AI Task Receipt recording on M Hash L2 is used for task verification.

Why Receipts Matter

AI Task Receipts are verifiable proof that a payment was made, a task was requested, and a result was returned. Any party — agent, user, auditor, or smart contract — can independently verify the full history on-chain.

Receipts are designed to be programmable: agents can reference them in subsequent transactions, dispute resolution contracts can read them, and compliance systems can audit them without requiring trusted intermediaries.

Contract Event

The AITaskReceiptCreated event is emitted on M Hash L2 when a receipt is finalized:

event AITaskReceiptCreated(
receiptIdbytes32
taskIdstring
useraddress
provideraddress
amountuint256
tokenaddress
paymentTxHashbytes32
timestampuint256
)
Developer Integration

Your Code + SKILL.md = An Aapp

A SKILL.md defines the agent-facing contract. Agents read it; services fulfill it. No proprietary SDK required.

// SKILL.md — AI Agent Service Contract { "skill": "wallet-risk-analysis", "version": "1.0", "provider": "MAGNE-Agent-001", "price": "0.01 mMHA", "chain": "mhash-l2-testnet", "chainId": 20250827, "paymentToken": "mMHA (ERC20)", "facilitator": "https://api.magne.ai/facilitator", "receiptContract": "0xa0Fe7AD300152eeDC4ca0c17514A1001D6afA408", "x402Compatibility": "under-validation" } // Agent flow: 1. POST /api/v1/wallet-risk → 402 + payment metadata 2. mMHA.transfer(recipient, amount) on M Hash L2 3. Facilitator.strictVerify(paymentTxHash, taskId, amount) 4. AI Task Receipt event → on-chain proof
View Full Demo Repo Developer Docs
Verification Status

Contract Deployment Verification

All verified items are supported by on-chain data. All pending items are actively being developed.

✓ Verified

  • ✓ M Hash L2 Testnet contract deployment
  • ✓ MockMHA test token contract
    0x13fadbC67B25870cF93DBa44f570f355dE456A2E
  • ✓ AITaskReceipt contract
    0xa0Fe7AD300152eeDC4ca0c17514A1001D6afA408
  • ✓ AI Task Receipt transaction reference
    tx: 0x07bc0ca32a037ab755b4516641fba45fe93ac33ed8864bdd9928d7ac152390fd (block 8603)
  • ✓ Public explorer links available

⏳ In Development

  • ⏳ Real HTTP 402 challenge and retry flow
  • ⏳ Browser wallet authorization
  • ⏳ Facilitator verify / settle workflow
  • ⏳ Stablecoin settlement support
  • ⏳ Formal x402 compatibility review
  • ⏳ Production security review

⚠ Security Notice: The currently disclosed testnet deployment is provided for developer demonstration purposes only. Production deployment will require controlled signing infrastructure, security review, contract administration procedures, and applicable compliance assessment.

Archived Deployment: The previously disclosed deployment at MockMHA 0x5FbDB2315678afecb367f032d93F642f64180aa3 / AITaskReceipt 0xe7f1725E7734CE288F8367e1Bb143E90bb3F0512 used a publicly-known test key. A new secure deployment has been completed at the addresses listed above. Do not use exposed-key deployments for any purpose beyond demonstration.

Ecosystem Position

Where MAGNE Fits in the Agent Payment Landscape

MAGNE is not competing with Coinbase, OKX, or Binance. It focuses on a specific layer: secure mobile authorization and verifiable AI Task Receipts on M Hash L2.

Protocol
Core Focus
MAGNE Alignment
x402 Foundation
Open payment protocol under Linux Foundation. Original design by Coinbase.
MAGNE references emerging x402-compatible patterns. No official integration.
OKX APP
Agent commercial workflow: intents, brokers, escrow, metering, disputes
Reference only — different scope
Binance x402
BNB Chain implementation for programmable payments
Reference only — different chain ecosystem
MAGNE Agent Pay
Secure mobile authorization + facilitator verification + AI Task Receipt on M Hash L2
Core focus — our differentiation
x402 Foundation: x402.org · Linux Foundation · Cloudflare x402

References to Coinbase, OKX, Binance, x402 Foundation, and BNB Chain are for ecosystem comparison only and do not imply partnership, endorsement, integration, or affiliation.

Ecosystem Positioning

Technical Research Reference

For technical research reference only. Does not indicate any partnership or official integration.

P0 — Open Protocol
x402 P0

HTTP 402-based open payment protocol for agent payments, initiated by Coinbase and transferred to the Linux Foundation in April 2026. Research reference for AI agent payment workflow; formal compatibility pending validation.

P1 — Industry Reference
B402 / Binance x402 P1

Binance's x402-compatible payment verification and submission service on supported chains. Included as a public industry reference; does not indicate MAGNE is integrated with Binance services.

Google AP2 P1

Google's agent-to-agent payment authorization protocol addressing liability, trust chains, and delegation for AI agents. Concept reference for mobile secure authorization.

Self-developed
MAGNE Agent Pay P0

Contracts and AI Task Receipt demonstration on M Hash L2 testnet. Focus on mobile secure authorization and on-chain receipt recording.

Positioning Note

This table is for reference only, describing publicly available technical directions and MAGNE Agent Pay's research positioning. It does not indicate that MAGNE.AI has any partnership, endorsement, or official integration with Coinbase, Linux Foundation, Binance, Google, or any third party, nor does it indicate completion of compatibility certification for any protocol or service.

Technical Validation Roadmap

What's Next

Subject to technical validation and security review.

✓ V0.1 — Done

Local Contract & AI Task Receipt Demo

Full payment flow verified on local Hardhat. ERC20 payments, Facilitator verification, on-chain receipts.

  • Smart contracts compile and deploy
  • Backend API with x402 response
  • Facilitator strict verify mode
  • AI Task Receipt events
⟡ V0.2 — Done

M Hash L2 Testnet Contract Deployment & Public Receipt Recording

M Hash L2 Testnet contract deployment, AI Task Receipt recording, and public explorer verification. Completed.

  • M Hash L2 testnet deployment
  • Real wallet MetaMask flow
  • Live explorer tx verification
  • Facilitator on testnet
○ V0.3 — In Development

Real HTTP 402, Browser Wallet Authorization & Facilitator Verification

Browser wallet authorization, HTTP 402 challenge/response flow, and Facilitator on-chain settlement verification. In active development.

  • MetaMask / browser wallet connection
  • HTTP 402 challenge and retry flow
  • Facilitator verify / settle workflow
  • End-to-end user authorization flow
○ V0.4 — Research

x402 Compatibility Testing & B402 / Binance x402 Service Model Research

Research and validation in progress.

  • x402 compatibility testing
  • B402 service model research
○ V0.5 — Planned

Stablecoin Settlement, MAGNE Phone Secure Authorization & Security Audit

Planned.

  • Stablecoin settlement support
  • MAGNE Phone secure authorization
  • Security audit preparation
○ V1.0 — Planned

Production Deployment After Security & Compliance Assessment

Subject to security audit, compliance assessment, and governance approval. Planned.

  • Security audit
  • MHA as settlement asset
  • Production Facilitator SLA
  • x402 spec compliance review

This roadmap describes product and technical validation directions and may be adjusted based on security review, technical results, protocol evolution, regulatory factors, and governance decisions. It does not constitute a commitment to feature launch timelines, partnerships, transaction support, or commercial outcomes.

⚠ Security Notice

The currently disclosed testnet deployment is provided for developer demonstration purposes only. Production deployment will require controlled signing infrastructure, security review, contract administration procedures, and applicable compliance assessment. This demonstration uses a temporary test key — do not use for production. MHA and mMHA are test tokens with no mainnet value.

Start Building.

Read the SKILL.md, fork the demo, and ship your first agent payment integration.

Fork the Demo Developer Docs Transparency Center