Do six UX patterns improve agentic AI user trust?

Victor Yocco's Smashing Magazine series proposes six UX patterns "that follow the functional lifecycle of an agentic interaction," grouped into three phases. "Pre-Action (Establishing Intent)" pairs the Intent Preview (also called Plan Summary) with the Autonomy Dial so "the user defines the plan and the agent's boundaries before anything happens." "In-Action (Providing Context)" pairs the Explainable Rationale with the Confidence Signal to show "the 'why' and 'how certain'" while the agent works. "Post-Action (Safety and Recovery)" pairs the Action Audit & Undo with the Escalation Pathway as "a safety net for errors or high-ambiguity moments."

Each pattern carries a stated psychological rationale. The Intent Preview "reduces cognitive load and eliminates surprise, giving users a moment to verify the agent truly understands their intent" before it acts. The Autonomy Dial, "a form of progressive authorization," treats trust as "a spectrum" rather than "a binary switch," letting users "match the system's behavior to their personal risk tolerance." The Explainable Rationale translates "system primitives into user-facing language to prevent deception," answering "why did it do that?" in terms grounded in "the user's own stated preferences and prior inputs" rather than "a technical log file." The Confidence Signal makes "the agent's internal state more legible" so users know "when to scrutinize a decision more closely." Action Audit & Undo is, in Yocco's framing, "the single most powerful mechanism for building user confidence," because "knowing that a mistake can be easily undone creates psychological safety." And because "errors are not a possibility; they are an inevitability," the Escalation Pathway has the agent escalate rather than guess when it is "uncertain about the user's intent or the best course of action."

The six-pattern set gives interaction-design vocabulary to claims the library already states more abstractly. The Autonomy Dial operationalizes the same spectrum that Can we design AI autonomy separately from capability? names as discrete roles, and the same target that How should AI agents balance autonomy with user oversight? describes as alignment across levels. The Intent Preview is a concrete mechanism for the requirement that How can users verify what an AI agent actually plans to do? — Yocco's pattern is one answer to how that verification should surface. Action Audit & Undo speaks directly to the finding that What makes people distrust AI agents they delegate to?: undo is Yocco's proposed fix for the exact irreversibility that study identifies as the trust-withdrawal trigger. And the Escalation Pathway stays on the intervention side of the categories that Does human-in-the-loop design actually enable effective oversight? argues builders over-emphasize, offering escalation rather than the knowledge or observation layers that primer says usually get skipped.

The excerpt names and motivates each pattern but reports no study, usability test, or deployment data showing any of the six actually changes user trust, task success, or error recovery in practice — the psychological underpinnings given for each pattern are stated as design rationale, not measured outcomes. It also does not say how Explainable Rationale or Confidence Signal should work when an agent's own confidence estimate is unreliable. As a design vocabulary for agentic UX, the six patterns read as a practical checklist; whether they hold up under use is a separate, open question.