Cobo Proposes "Smart Contracted Permission + MPC" Framework: Building a Controllable Trust Layer for the Agent Economy

BlockBeats News, April 21st, Cobo AI Growth Lead Brad Bao shared at the offline event "Decoding Web 4.0: When AI Agent Takes Over On-Chain Authority" that Cobo is restructuring the fund control logic of the Agent through a structured permission system. In its latest release of the Cobo Agentic Wallet, the relationship between the user and the Agent is no longer a simple wallet authorization but is defined by an executable "Pact" to establish permission boundaries.

This "Pact" includes at least four core elements: Intent, Execution Plan, Policies, and Completion Conditions, transforming the original "all or nothing" permission control into a fine-grained governance mechanism that is negotiable, auditable, and revocable.

At the execution level, Cobo combines MPC (Multi-Party Computation) technology to ensure that even in cases of Agent misbehavior, system attacks, or risk of mnemonic phrase injection, the funds remain controlled by the user, and no single entity can bypass the constraints to independently transfer assets. Brad Bao emphasized that under this architecture, the Agent has enough freedom to carry out complex tasks, but its behavior is always confined within verifiable boundaries.

He stated that this system essentially establishes a layer of "trust and insurance infrastructure" in the Agent economy, achieving controlled empowerment of Agent behavior through a combination of "contractual permissions + MPC". This ultimately enhances automation efficiency while safeguarding user asset security and result determinism.