From "Can Pay" to "Getting Things Done": X-Agent Is Building an Onchain Personal AI Assistant

Odaily reports that X-Agent has published its latest technical article, "From Agent Payments to an Onchain Personal Assistant," unveiling its next major product direction: moving beyond Agent Payments toward an Onchain Personal Assistant capable of understanding user intent, obtaining authorization, executing onchain operations, and continuously tracking results.

Unlike simply giving an AI Agent a "wallet," X-Agent aims to solve a more critical problem: how to enable an Agent to safely complete an onchain task on behalf of the user from start to finish.

For example, when a renewal bill arrives, the AI assistant can interpret the bill, verify the payee and amount, generate a payment plan, and complete the transaction after the user authorizes it. Even if a timeout or interruption occurs during the process, it should resume the original task and verify the payment status, rather than blindly repeating the payment. After the payment is completed, the Agent also needs to continue confirming whether the service the user purchased is actually delivered.

To achieve this goal, X-Agent plans to further connect three major modules—Agent ID, Agentic Email, and Agentic Wallet—which respectively handle identity and accountability, task and communication entry points, and controlled onchain fund execution. Through mechanisms such as payment pre-computation, one-time authorization, budget control, recovery, and reconciliation, X-Agent aims to establish clear security boundaries for autonomous Agent execution.

Its core logic can be summarized as:

User intent → Explicit authorization → Onchain execution → Result verification

X-Agent believes that a true onchain AI assistant is not merely one that "initiates a transaction on behalf of the user," but one that can continuously complete tasks under controlled permissions and take responsibility for the final outcome. This also means that X-Agent is evolving from Agent payment infrastructure to further exploring a persistent onchain AI assistant for individual users.