How are you using a Desktop or Browser AI agent daily?

There are a lot of tools in the market, Claude Code, HeyClicky, GPT-6 Astra, Meta Muse any other, that can "see" your screen and act on your behalf. Agentic AI, AGI terms are trending in market and creating havoc among emerging developers and freshers. I'm curious to hear from developers and power users: 1. Are any of you using a screen-aware AI agent in your daily workflow? 2. If yes, which one? How is it, and what are the pros and cons? 3. If no, is it because of privacy concerns, latency, or just a lack of actual utility? Trying to separate the hype from the reality of agentic desktop workflows.