Proof Generation

Ungate Wukong provides three fundamental cryptographic proofs to ensure the integrity of autonomous agents:

1. Proof of Autonomy (PoA)

Verifies that an agent is executing independently without hidden human intervention or external manipulation.

• Attested by the TEE or virtual enclave: Confirms that execution was not externally altered.

• Logs signed with a cryptographic key: Ensures the integrity of execution data.

• Decentralized verification: Prevents a single party from faking agent autonomy.

2. Proof of Sovereignty (PoS)

Ensures that an agent has exclusive control over its own wallets, private keys, and decision-making logic.

• Prevents unauthorized developer access: Developers cannot alter execution once deployed.

• Guarantees on-chain wallet sovereignty: The agent alone holds and manages its assets.

• Mitigates legal and financial liability risks: Developers cannot be held responsible for agent-driven financial transactions.

3. Proof of Execution Integrity (PoEI)

Verifies that an agent has executed the correct logic, interacted with the intended data sources, and used the models it claimed to use.

• Ensures AI model integrity: Prevents agents from calling unauthorized external models.

• Protects against data manipulation: Guarantees that transaction execution is as expected.

• Reduces attack vectors: Agents cannot misrepresent execution results.

Proof Generation Process

1. The agent runs in a secure enclave.

2. The execution process is cryptographically signed.

3. Proofs are recorded on decentralized storage (EigenDA, FileCoin).

4. EigenLayer validators verify and attest to execution integrity.

This proof-based framework ensures that every agent in Ungate Wukong is cryptographically verifiable and provably autonomous.