world_ai_protocol

Report Abuse

Basic Information

WORLD AI Protocol is a chain-agnostic on-chain delegation framework designed to let external agents, including AI agents or service accounts, act autonomously on behalf of users (principals) while preserving security and user control. The repository describes a permissioned system that records per-principal, per-agent, per-function authorizations with time bounds and usage limits, requires agent consent via signature, and supports batch operations and event logging. It includes reference implementations and guidance for EVM and a Move-based Sui implementation, and documents core mappings, structs, and the lifecycle for creating, checking, updating, and revoking authorizations. The main purpose is to provide developers and dApp teams a reusable protocol to enable governed automation across Web3 environments without exposing private keys or granting unlimited authority.

Links

App Details

Features
This protocol implements function-level permissions so agents are authorized by specific function selectors. Each agent is limited to a single principal at a time and authorizations include start and end timestamps for time-bound access. Usage allowances limit how many times an agent can invoke an authorized function and decrement automatically on each call. Authorizations are automatically removed when usage is depleted and principals may dynamically update or revoke permissions at any time. Agent consent is enforced via signatures. Batch authorizations allow multiple changes in a single transaction for efficiency. The design is chain-agnostic with reference EVM and Sui/Move implementations and emits events for real-time auditing of authorization changes.
Use Cases
WORLD AI Protocol enables secure, granular automation for Web3 use cases by decoupling repetitive actions from a principal"s private keys while preserving fine-grained control. It supports autonomous gaming actions, DeFi yield strategies, on-chain workflow orchestration, and multi-chain agent coordination by restricting agents to approved functions, time windows, and usage counts. Principals retain revocation and update power and receive auditability through emitted events. Batch operations reduce transaction overhead for large-scale updates. The README also outlines a roadmap for role-based permissions, gasless meta-transactions, cross-chain expansion, and multi-call conveniences to further simplify AI-driven or script-based automation in decentralized applications.

Please fill the required fields*