How do I install Agent Pay?

Copy the stdio config for Cursor, Claude Desktop, Claude Code, or VS Code on this page. It uses the publisher's install, not a PlugMCP proxy.

Does Agent Pay need an API key?

The publisher lists these environment variables: AGENT_PAY_BURNER_SEED, AGENT_PAY_AGENT_KEY.

What transport does Agent Pay use?

Agent Pay runs locally over stdio.

When was Agent Pay last checked?

Last checked 2026-11-08. x402 payments in USDC on Base inside an owner-signed, on-chain budget the model can't change

Agent Pay MCP server

Power your agent

Add this server

The publisher's own command. Not a PlugMCP endpoint.

Add to Cursor Add to VS Code Run on PlugMCP

Host a copy with our keys and logs.

Local command

npx -y @deepfirstsearch/[email protected]

Credentials from the publisher

AGENT_PAY_BURNER_SEED — 32-byte hex secret the per-merchant payer addresses are derived from. Never your owner key.
AGENT_PAY_AGENT_KEY — The budget's agent key (needed when a vault is configured).

Claude Code has no install link for a local command. Copy the command below and run it in a terminal.

shClaude Code

1AGENT_PAY_BURNER_SEED=<secret> AGENT_PAY_AGENT_KEY=<secret> claude mcp add com.deepfirstsearch-agent-pay-mcp -- npx -y @deepfirstsearch/agent-pay-mcp@0.2.1

Run in a terminal.

jsonClaude Desktop

1{2  "mcpServers": {3    "com.deepfirstsearch-agent-pay-mcp": {4      "command": "npx",5      "args": [6        "-y",7        "@deepfirstsearch/[email protected]"8      ],9      "env": {10        "AGENT_PAY_BURNER_SEED": "<secret>",11        "AGENT_PAY_AGENT_KEY": "<secret>"12      }13    }14  }15}

claude_desktop_config.json

Local · npmPublisher installStep-by-step: add this in Claude

Remote URL or a command on your machine? How the two installs differ

Run on PlugMCP

Host a copy with our keys and logs.

A new server you create. It is not this listing's official endpoint. Use the publisher install when you want their URL and their credentials.

Launch

Server facts

Local

Transport: stdio
Version: 0.2.1

Tools

No tool list yet. Remote servers get one after the first health check.

FAQ

Questions

Deep First Search: safe x402 payments for AI agents

Safe rails for the agent economy. Search deep. Reveal only what you choose.

AI agents now pay for things over x402. Every one of those payments is public, and any agent can be talked into paying the wrong party. Deep First Search makes the model propose payments while the owner's signed policy decides.

Try it in 10 seconds

No keys, no wallet, no chain: a local x402 merchant plays honest and hostile, and you watch the agent pay for data and refuse four attacks.

npx @deepfirstsearch/agent-pay demo

Pick your path

You use	Install	Guide
Claude Desktop, Claude Code, Cursor, OpenClaw, any MCP client	`npx @deepfirstsearch/agent-pay-mcp ./config.json`	integrations/mcp
Vercel AI SDK	`npm i @deepfirstsearch/agent-pay-ai-sdk @deepfirstsearch/agent-pay ai zod viem`	integrations/ai-sdk
LangChain.js / LangGraph	`npm i @deepfirstsearch/agent-pay-langchain @deepfirstsearch/agent-pay @langchain/core @langchain/langgraph viem`	integrations/langchain
Coinbase AgentKit	`npm i @deepfirstsearch/agent-pay-agentkit @deepfirstsearch/agent-pay @coinbase/agentkit [email protected] zod@3`	integrations/agentkit
Your own agent, any wallet	`npm i @deepfirstsearch/agent-pay viem`	sdk
A Turnkey-held key	`npm i @deepfirstsearch/agent-pay-turnkey @deepfirstsearch/agent-pay @turnkey/sdk-server viem`	Turnkey
A Privy server wallet	`npm i @deepfirstsearch/agent-pay-privy @deepfirstsearch/agent-pay @privy-io/node viem`	Privy
A Coinbase CDP Server Wallet	`npm i @deepfirstsearch/agent-pay-cdp @deepfirstsearch/agent-pay @coinbase/cdp-sdk viem`	CDP
An Openfort backend wallet	`npm i @deepfirstsearch/agent-pay-openfort @deepfirstsearch/agent-pay @openfort/openfort-node viem`

The integrations take the SDK as a peer dependency, so your app always uses a single copy of it. Step-by-step guides for 40 x402 merchants, each tested with a real payment on Base mainnet, are in docs/integrations.

Full developer guide: deepfirstsearch.com/developers · building a wallet, framework or platform? Partners.

What's in this repository

Component	What it does	Status
MCP server (`integrations/mcp/`)	Model Context Protocol server: Claude Desktop, Claude Code, Cursor and any MCP client get `paid_fetch`, `list_merchants` and `budget_status` tools that pay x402 APIs inside the owner's budget; x402 v1 merchants opt in per merchant (`x402Versions`), and an unreadable receipt is confirmed on-chain instead of resent (`confirmOnChain`, on by default)	✅ `npx @deepfirstsearch/agent-pay-mcp ./config.json` (0.2.1) · 24 tests · tested with a real Claude session on Base Sepolia
Agent Safe (`contracts/`)	On-chain vault: owner-signed, timelocked, per-merchant budgets; the agent key can spend but never widen	✅ 98 tests (fuzzing, invariants, properties, regressions, Base mainnet fork with real USDC), two internal reviews · live on Base mainnet (beta)
agent-pay SDK (`sdk/`)	x402 v2 client (v1 opt-in per merchant) with a policy engine outside the model, prompt-injection guards, stable refusal and block codes, never a second signature while the first might have paid, per-merchant payers, owner CLI	✅ 293 tests (283 run offline; 10 need a Base mainnet fork or RPC) incl. on-chain loops and the official x402 facilitator · `npx @deepfirstsearch/agent-pay owner …`
Framework tools (`integrations/`)	Vercel AI SDK `paidFetchTool`, LangChain/LangGraph `createPaidFetchTool`, Coinbase AgentKit `agentPayActionProvider`; wallet payers `turnkeyPayer`, `privyPayer`, `cdpPayer`, `openfortPayer`; `order-guard` for agents that place real orders	✅ on npm
$DEPTH (`contracts/`)	1B fixed supply, no mint function, no owner; fees burned through a fee jar + firepit	✅ Implemented · not launched
Website (`web/`)	Site, whitepaper, tokenomics, partner kit, legal pages; strict CSP, no cookies; only third party is cookieless Cloudflare Web Analytics	✅

MCP server: let Claude or Cursor pay x402 APIs safely

@deepfirstsearch/agent-pay-mcp is a stdio MCP server. The model gets three tools and nothing that can move money on its own terms:

MCP tool	What it does
`paid_fetch(url, method?, body?, contentType?)`	Fetches a URL; if it answers `402 Payment Required`, pays in USDC on Base only when the merchant, price and budget match your config
`list_merchants()`	Merchants the agent may pay, their prices and the remaining budget
`budget_status()`	Remaining budget per merchant, plan window, vault balance

{ "mcpServers": { "agent-pay": {
    "command": "npx", "args": ["-y", "@deepfirstsearch/agent-pay-mcp", "/path/config.json"],
    "env": { "AGENT_PAY_AGENT_KEY": "0x…", "AGENT_PAY_BURNER_SEED": "0x…" } } } }

Payees, prices and limits come from the config file and the owner-signed on-chain budget, never from the model or a 402 response, so a prompt-injected agent can't overspend or pay an attacker. Setup, config reference and a 5-minute Base Sepolia demo: integrations/mcp.

Status: pre-launch. Live on Base Sepolia, with a real x402 payment settled (see deployments); unaudited beta on Base mainnet since 2026-10-06 (small amounts, our own funds first, then invited design partners) while we arrange an independent audit; the public launch comes after the audit. No token exists.

Develop on this repository

# Contracts (needs Foundry: curl -L https://foundry.paradigm.xyz | bash && foundryup)
git submodule update --init --recursive
cd contracts && forge test

# SDK
cd ../sdk && npm ci && npm test && npm run demo

# Website
cd ../web && npm ci && npm run dev

Documents


Whitepaper	Formal design, privacy model, risks (EN; structured for MiCA Annex I)
Tokenomics	Supply, allocation, Firepit burn, vesting (EN)
Decisions	ADRs: ticker, chain, burn, vesting, stand… (ES)
Deployments	Contract addresses on Base mainnet (unaudited beta) and Base Sepolia, plus the live demo vault
Audit scope	Scope, nSLOC, roles, known issues and questions for reviewers
Internal pre-audit, Oct 2026	Every finding of our latest internal review, fixed (with its test) or acknowledged (with the reason)
Airdrop tool	Builds the airdrop Merkle tree and proofs; tested against the contract
Mainnet runbook	Safes, rehearsal, deploy, checks, guarded launch, monitoring
Security assessment	Pentest findings, threat model, evidence (ES)
Research	Market pain, successes, failures, token models, launch mechanics (ES)

Key decisions

Ticker: $DEPTH.
Chain: Base.
Founder: 12%, nothing for 12 months, then 36 months linear, on-chain; the vesting contract cannot be transferred and the beneficiary is a public Safe.
Burn: burn-to-claim fee jar (no swaps, oracles or admins), fed only by Agent Safe, SDK and inference fees, never by privacy-pool fees.
No conference booth until there are users; hackathons instead.
No own privacy pool: integrate a third-party compliant one (ADR-011). Sales geo-blocked for the US and Argentina (ADR-012).

Using it? Tell us

There's no telemetry in the packages, so we only know who uses them if you say so. If your agent pays with Agent Safe, even once on testnet, tell us in this discussion: what it pays for, what got refused that shouldn't have, and what's missing. We'll test your sellers live and write a verified guide for any that's missing.

Security

See SECURITY.md to report vulnerabilities.

Contact

Built by Nicolas Tursi. General contact: [email protected]. Security reports: see SECURITY.md or write to [email protected].

License

MIT (contracts and SDK).

Add this server

Host a copy with our keys and logs.

Server facts

Tools

Questions

How do I install Agent Pay?

Does Agent Pay need an API key?

What transport does Agent Pay use?

When was Agent Pay last checked?

Documentation

Deep First Search: safe x402 payments for AI agents

Try it in 10 seconds

Pick your path

What's in this repository

MCP server: let Claude or Cursor pay x402 APIs safely

Develop on this repository

Documents

Key decisions

Using it? Tell us

Security

Contact

License