Model Context Protocol server
Give any MCP-capable agent Paymod payment tools without exposing wallet signing authority.
url: mcp.paymod.xyzEach Paymod Agent Wallet has its own isolated balance and authority. A policy mistake or a compromised credential for one agent should not expose another agent's funds. Policy limits what the agent is allowed to spend. Wallet isolation limits what it can physically reach.
When an agent encounters an x402-protected resource, Paymod evaluates the payment against its wallet authority, pays when permitted and returns the resource, without ever handing the agent unrestricted signing power.
approver ping sent
Deterministic by design. Same intent, same active policy, same decision.
Give any MCP-capable agent Paymod payment tools without exposing wallet signing authority.
url: mcp.paymod.xyzPrewritten skills for pay, quote, refund and balance checks.
skills/pay-x402.mdSubmit a financial intent, receive the authorization decision and track execution through one typed client.
await paymod.pay({...})Sign a request, get a decision. Webhooks stream every state change back.
POST /v1/intentsPaymod Code is a pay-as-you-go coding agent for VS Code and the terminal that reads, edits, runs and tests your real projects using multiple model providers, with transparent usage metering through Paymod.
Every intent, decision, approval and execution result can be traced back to the authority that produced it.
| Entry | Time | Agent | Intent | USDC | Decision | Authority |
|---|---|---|---|---|---|---|
| #10,482 | 09:41:02 | ops-agent | transfer · base | 180.00 | APPROVED | @mara |
| #10,481 | 09:40:55 | research-agent | x402 · serpstack | 0.42 | ALLOW | policy v12 |
| #10,480 | 09:38:11 | growth-agent | x402 · ads.exchange | 12.50 | ALLOW | policy v12 |
| #10,479 | 09:36:47 | unknown-agent | transfer · anon | 2,400.00 | DENY | no permission |
| #10,478 | 09:31:20 | research-agent | x402 · openrouter | 0.08 | ALLOW | policy v12 |
| #10,477 | 09:22:04 | ops-agent | refund · stellar | 40.00 | APPROVED | @dre |
Each agent gets its own wallet with an isolated balance and authority. A policy mistake or a compromised credential for one agent cannot reach another agent's funds.
Rules are deterministic: the same intent evaluated against the same active policy always produces the same decision, checked before any funds move.
Depends on the chain. Paymod is the policy, approval and audit layer every agent payment passes through, regardless of what runs underneath. For EVM chains that layer sits in front of Circle's Developer-Controlled Wallets; for other chains, including Stellar and Solana, Paymod builds and runs the wallet infrastructure itself. Either way, the agent never receives unrestricted signing power, only a revocable wallet credential Paymod's policy engine checks on every request.
When an agent hits an x402-protected resource, Paymod evaluates the payment against that agent's wallet authority, pays when permitted and returns the resource.
Yes. Every intent, decision, approval and execution result is written to a ledger that traces back to the authority that produced it.
Through an MCP server, drop-in agent skills, a typed SDK or a plain REST API, whichever fits your agent's runtime.