How Agents Use Wallets

The wallet is what makes an agent onchain. It is also the single biggest risk surface: an agent with unconstrained signing power can lose everything it holds. The agent wallet stack (embedded wallets, secure enclaves, policy engines) is what makes autonomous execution viable for real value.

• ▸Keys are held by infrastructure (secure enclaves, MPC, or custodial APIs) rather than in the agent's prompt or code.
• ▸The agent requests signatures through an API; a policy engine evaluates each request against rules (spend limits, allowlisted contracts, time windows).
• ▸High-risk actions can require human approval before signing.
• ▸Session keys and scoped permissions let deployers grant narrow, revocable capabilities.

• •Key custody (enclave, MPC, or managed)
• •Policy engine (limits, allowlists, approvals)
• •Signing API
• •Audit logs
• •Revocation and kill switches

• ▸A treasury agent with a $500 daily spend limit and a three-contract allowlist.
• ▸A trading agent using a policy-controlled signer that blocks withdrawals to unknown addresses.
• ▸A payments agent using x402 to pay per-request for API access in USDC.

• ⚠Key leakage through logs, prompts, or compromised infrastructure.
• ⚠Policy gaps: an agent can do anything its policies don't forbid.
• ⚠Social engineering and prompt injection aimed at triggering transfers.
• ⚠Custody questions: who actually controls the agent's funds?

• What Are Onchain Agents?
• How Agents Interact With Smart Contracts