AESP FAQs

Question 1

What is AESP and its primary purpose?

Accepted Answer

AESP (Agent Economic Sovereignty Protocol) is a TypeScript SDK designed to enable AI agents to conduct economic transactions while ensuring explicit human oversight and control. Its primary purpose is to maintain human sovereignty over AI's financial actions.

Question 2

What types of economic transactions can AI agents perform using AESP?

Accepted Answer

Through AESP's MCP tools, AI agents can perform various economic actions, including checking vault balances, depositing funds, redeeming vault shares, transferring allocations between accounts, and checking authorization statuses, all within the bounds of human-defined policies.

Question 3

Can AESP integrate with existing AI agent frameworks like Claude?

Accepted Answer

Yes, AESP is designed for practical integration with AI agent frameworks. It provides clear setup instructions for adding it to Claude Desktop and Claude Code as an MCP server, allowing agents to access its economic tools seamlessly.

Question 4

How does AESP ensure human control over AI agent spending?

Accepted Answer

AESP ensures human control through 'policy-gated execution,' which uses customizable rules for budgets, spending limits, and address allowlists. It also provides a 'human override path' that escalates risky actions for manual review and approval, preventing unauthorized or unintended transactions.

Question 5

What are the key technical features AESP offers developers?

Accepted Answer

For developers, AESP offers a modular TypeScript SDK with subpath exports for targeted imports, pre-built MCP (Multi-Agent Communication Protocol) tools for common vault operations, and EIP-712 signed intent for verifiable on-chain settlement guarantees. It also includes policy engines, identity management, and negotiation state machines.

AESP

AESP

主な機能

ユースケース

主な機能

ユースケース