Stats
Actions
Tags
From agentpay
Cap and receipt an AI agent's spending. Open a budget-capped AgentPay session so paid tool/API calls stay under a hard USDC limit and each produces a verifiable receipt + a running ledger. Use when enforcing a spend cap across calls, tracking what an agent spent, or producing an audit trail for autonomous payments. Pairs with the agentpay-route skill.
How this skill is triggered — by the user, by Claude, or both
Slash command
/agentpay:agentpay-sessionThe summary Claude sees in its skill listing — used to decide when to auto-load this skill
Once the agent has chosen a paid tool (see the **agentpay-route** skill), wrap the payment in
Once the agent has chosen a paid tool (see the agentpay-route skill), wrap the payment in an AgentPay session so spend stays under a hard cap and every call leaves a verifiable receipt. This is the spend-governance layer: budget enforcement + receipts + a running ledger across calls, multi-chain (USDC on Base or Stellar), peer-to-peer (AgentPay never holds funds).
pip install agentpay-x402
from agentpay import Session, AgentWallet
s = Session(AgentWallet(secret_key="S...", base_key="0x..."), max_spend="0.05")
s.remaining_usd() # budget left, before committing to a call
r = s.call("<provider-url-or-tool>", {...}) # pays within the cap; raises if over budget
print(s.spending_summary()) # receipt: each call's cost + tx + chain + remaining
Zero-setup variant for the free tools / trying it: from agentpay import quickstart; s = quickstart(max_spend="0.10").
prefer_chain="stellar"
to override). Free ($0) tools never settle on-chain and need no wallet.Home: https://agentpay.tools · SDK: pip install agentpay-x402
Creates, edits, and optimizes skills for Claude Code, including drafting, evaluating with test prompts, iterating on performance, and improving skill descriptions for better triggering accuracy.
npx claudepluginhub romudille-bit/agentpay --plugin agentpay