---
title: "Comment by mandely on \"20 Emails, 20 Clicks: Gemini Spark Cleaned My Inbox and Proved Why \"Agentic\" Has Limits\" | Hashnode"
description: "The hesitation around destructive actions feels less like a capability gap and more like a trust calibration problem. That's probably one of the biggest UX chal"
url: "https://hashnode.com/posts/20-emails-20-clicks-gemini-spark-cleaned-my-inbox-and-proved-why-agentic-has-limits/6a419437ba11b22757f450f2/comment/6a46373b946c5a9d44713c1a"
---

[Discussion](https://hashnode.com/posts/20-emails-20-clicks-gemini-spark-cleaned-my-inbox-and-proved-why-agentic-has-limits/6a419437ba11b22757f450f2)

Comment [20 Emails, 20 Clicks: Gemini Spark Cleaned My Inbox and Proved Why "Agentic" Has Limits](https://hashnode.com/posts/20-emails-20-clicks-gemini-spark-cleaned-my-inbox-and-proved-why-agentic-has-limits/6a419437ba11b22757f450f2)

[mandely](https://hashnode.com/@mandely)

[Jul 2](https://hashnode.com/posts/20-emails-20-clicks-gemini-spark-cleaned-my-inbox-and-proved-why-agentic-has-limits/6a419437ba11b22757f450f2/comment/6a46373b946c5a9d44713c1a)

The hesitation around destructive actions feels less like a capability gap and more like a trust calibration problem. That's probably one of the biggest UX challenges for AI agents going forward.