agentpay-mcp 4.1.0 → 4.1.1

This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
Files changed (2) hide show
  1. package/README.md +37 -0
  2. package/package.json +2 -2
package/README.md CHANGED
@@ -584,6 +584,43 @@ agentpay-mcp is the **open-source governance layer** on top of this infrastructu
584
584
 
585
585
  ---
586
586
 
587
+ ## OpenAI Delegated Payment Spec Compatibility
588
+
589
+ OpenAI has published a **Delegated Payment Spec** that defines how AI agents handle payments on behalf of users: a scoped token with an allowance cap, compatible with Stripe Scoped Payment Tokens (SPTs). This is the precursor architecture to native payment tooling in OpenAI's Agents SDK.
590
+
591
+ **agentpay-mcp's spending cap model directly aligns with the Delegated Payment Spec pattern:**
592
+
593
+ | Delegated Payment Spec Concept | agentpay-mcp Implementation |
594
+ |---|---|
595
+ | **Scoped token** — agent receives limited-scope credential | `AGENT_PRIVATE_KEY` — agent signs within smart contract constraints, cannot exceed scope |
596
+ | **Allowance cap** — maximum the agent can spend | `set_spend_policy` — per-tx and daily caps enforced on-chain by AgentAccountV2 |
597
+ | **Human approval** — user delegates, then agent executes | `queue_approval` — transactions above threshold require explicit human sign-off |
598
+ | **Audit trail** — all delegated spend is logged | `get_transaction_history` — immutable on-chain event log per transaction |
599
+ | **Revocation** — user can revoke delegation at any time | Spend policy updates are instant; wallet owner can freeze the agent key |
600
+
601
+ The core pattern is identical: **human approves a budget → agent executes within that budget → all activity is auditable**. The difference is settlement layer: OpenAI's spec targets Stripe SPTs (fiat rails), while agentpay-mcp settles on-chain (USDC/ETH on Base, Arbitrum, and 8 other EVM chains).
602
+
603
+ For developers building OpenAI Agents SDK workflows that need on-chain settlement or multi-rail payment execution, agentpay-mcp serves as the MCP payment tool that implements the Delegated Payment Spec pattern with on-chain enforcement rather than application-level trust.
604
+
605
+ ```json
606
+ {
607
+ "mcpServers": {
608
+ "agentpay": {
609
+ "command": "npx",
610
+ "args": ["agentpay-mcp"],
611
+ "env": {
612
+ "AGENT_PRIVATE_KEY": "0x...",
613
+ "AGENT_WALLET_ADDRESS": "0x..."
614
+ }
615
+ }
616
+ }
617
+ }
618
+ ```
619
+
620
+ Add this MCP server to any OpenAI Agents SDK workflow via MCP bridge. The agent gets `x402_pay`, `check_budget`, and `set_spend_policy` — the same scoped-token + allowance-cap pattern, enforced by smart contract.
621
+
622
+ ---
623
+
587
624
  ## EU AI Act Compliance
588
625
 
589
626
  **Enforcement deadline: August 2, 2026.** AI systems that execute or facilitate financial transactions are classified as **high-risk** under EU AI Act Annex III. High-risk classification requires:
package/package.json CHANGED
@@ -1,8 +1,8 @@
1
1
  {
2
2
  "name": "agentpay-mcp",
3
- "version": "4.1.0",
3
+ "version": "4.1.1",
4
4
  "mcpName": "io.github.up2itnow0822/agentpay",
5
- "description": "AgentPay MCP Server \u2014 Non-custodial x402 payment layer for AI agents. Multi-chain wallets, spending limits, and machine-to-machine payments. Patent Pending.",
5
+ "description": "AgentPay MCP Server Non-custodial x402 payment layer for AI agents. Multi-chain wallets, spending limits, and machine-to-machine payments. Patent Pending.",
6
6
  "main": "dist/index.js",
7
7
  "types": "dist/index.d.ts",
8
8
  "bin": {