Suggest / Confirm / Execute
Suggest / confirm / execute is an AI interface design pattern that offers three discrete, user-selectable autonomy modes for an agent: suggest-only (agent proposes, user does the work), confirm-each-step (agent proposes and executes one step at a time with approval), and end-to-end execute (agent runs the whole workflow unattended). This UX pattern makes autonomy a visible, adjustable slider rather than an implicit, hidden property of the system. The current mode is always displayed, switching is one click, and users can change their minds mid-session. The pattern prevents the common failure of implicit autonomy, where users discover the agent is acting on its own only after seeing unexpected results, and gives teams a shared vocabulary for agent configuration.
Use Case
Essential for AI coding agents, workflow automation, and any interface where users need explicit control over how independently an AI acts on their behalf.
Examples in Wild
Use this pattern in your project
Copy this prompt to generate a production-ready implementation in Cursor, Claude Code, Lovable, or any AI coding agent.
Generate a production-ready implementation of the "Suggest / Confirm / Execute" AI interface design pattern.
Pattern Description:Draft response to vendor email
Update pricing row in CRM
Schedule outbound send
Irreversible if sent. This step is never auto-approved.
Related Patterns
Get new patterns by email
Weekly AI interface UX notes and resources on Substack, no spam, unsubscribe anytime.