Three core operations
| Operation | What it does |
|---|---|
| Send | Agents initiate USDC transfers without per-tx human approval — for API fees, bookings, microtransactions, multi-agent settlement. |
| Receive | Each agent wallet holds a dedicated address. Webhook events fire on every transfer. |
| Verify | Cryptographically confirm a payment before fulfilling any request. Gate agentic actions on confirmed USDC receipt — no trust required. |
Programmable safety layer
- Per-transaction limits — Cap the maximum USDC value of any single agent-initiated transfer.
- Daily spending caps — Rolling 24-hour limit, automatic reset, alerts as thresholds approach.
- Allowlist-only destinations — Restrict sends to pre-approved wallet addresses; agents cannot route outside the allowlist.
- Immutable audit trail — Every send/receive/verify call is logged on-chain and queryable via API.
Supported agent frameworks
Claude (Anthropic)
Use
wallet.asTool() to expose the wallet as an Anthropic tool.ChatGPT / GPT (OpenAI)
Compatible with the OpenAI tool-calling interface.
OpenClaw (OpenAi)
Native integration with OpenAi’s agent stack.
Custom agents
Works with any AI SDK — Anthropic, OpenAI, LangChain, or your own runtime.
Setup flow
Expose as a tool
wallet.asTool() passes it to the AI SDK; the agent can now call send/receive/verify.Quickstart
Use cases
Consumer
- Travel booking — AI books and pays for flights
- Subscription management — cancel, renew, switch providers
- AI shopping checkouts
- AI paying its own tools as it works
Developer / enterprise
- AI API metering
- x402 protocol payments
- Multi-agent settlement
- Data marketplace access
- Compute procurement (GPU time, cloud functions)
- Agent-to-agent commerce
Custom solutions
Off-the-shelf wallets cover most agent use cases. For consumer products, enterprise rollouts, or regulated agent platforms, REM builds the policy and oversight layer with you.Policy frameworks
Custom rule engines — categorical limits, merchant allowlists, time windows, multi-sig escalation, and human-in-the-loop fallbacks.
Audit dashboards
Operator-facing dashboards for monitoring agent spend, anomaly detection, and post-hoc forensic review.
Custom agent SDKs
Wallet primitives wrapped into your own SDK — branded, scoped to your agent runtime, and pre-integrated with your auth.
Regulated agent platforms
KYC-bound agent identities, supervised modes, and reporting pipelines for regulated agent operators.
Build with us
Tell us about the agent, the policy, and the volume. We’ll scope a custom wallet stack.
Wallet API reference
Endpoint specs for creating, querying, and managing agent wallets.

