---
title: "LinkedIn Ad - Early Release Claude for Task Automatio"
url: https://stacklist.com/card/a87911e4-0f2f-4885-8311-b64d7a839036
source_url: "https://www.linkedin.com/posts/claude_put-claude-to-work-on-your-computer-from-activity-7441961021650628608-r6uy?utm_source=share&utm_medium=member_ios&rcm=ACoAABavkckBv0rX_6ZwwcseyCd2I7U0CBYZJ98"
stack: https://stacklist.com/stack/b2a8ee22-cf1c-41a5-8408-a9c694e7de78
summary: "Claude now enables AI agents to control your computer directly, automating tasks like opening apps, navigating browsers, and filling spreadsheets through a research preview available on Pro and Max plans for macOS. The feature prioritizes API integrations with Slack and Calendar before falling back to screen control, representing a shift toward reliable agent execution over simple desktop automation."
tags: "claude, ai-agents, computer-automation, research-preview, productivity, anthropic"
key_entities: "Claude (technology), Anthropic (organization), Claude Cowork (technology), Claude Code (technology), Slack (technology), Calendar (technology), AI agents (concept), desktop-automation (concept), Piotr Szczepanik (person), Vincent Nwaka (person), Anu Verma (person), Tolani Omitokun (person), Will Tygart (person)"
classification: "reference"
content_hash: "sha256:564ec6dd00ebc8468ed6f5f922ab1202217f06ca41a53378f5f25a42669cb6f7"
acp_version: "0.2"
token_counts_approximate: 1552
visibility: public
agent_accessible: true
status: "final"
---

# LinkedIn Ad - Early Release Claude for Task Automatio

Put Claude to work on your computer, from anywhere Claude 2,014,782 followers 1mo Edited Report this post Now in research preview: You can enable Claude to use your computer to complete tasks in Claude Cowork and Claude Code. It opens your apps, navigates your browser, fills in spreadsheets—anything you'd do sitting at your desk. Claude uses your connected apps first: Slack, Calendar, and other integrations. When there's no connector for the tool you need, it asks for your permission to open the app on your screen directly. Assign a task from your phone, turn your attention to something else, and come back to finished work on your computer. The conversation picks up where it left off—tell Claude once to scan your email every morning or pull a report every Friday, and it handles it from there. It won't always work perfectly, and complex tasks could need a second try. We're sharing it early because we want to learn where it works and where it falls short. Available on Pro and Max, macOS only. Update your desktop app and pair with mobile to try: https://lnkd.in/gFEAnNZm …more 4,932 226 Comments Like Comment Share Copy LinkedIn Facebook X Piotr Szczepanik 1mo Report this comment The last paragraph is why I trust Anthropic more than most AI companies. "It won't always work perfectly" - posted on their own product announcement. Try finding that line in a competitor's launch post. The real unlock here isn't "AI uses your computer." OpenClaw does that. Perplexity Personal Computer does that. The difference is the integration stack - Slack, Calendar, and MCP connectors first, screen control as fallback. That's the right architecture. Screen scraping is fragile. API connections are reliable. Use the reliable path when you can, fall back to screen when you must. "Assign a task from your phone, come back to finished work" sounds like magic until you realize it's the same promise every AI agent is making right now. The question is execution. And Anthropic shipping this as research preview instead of "revolutionary product launch" tells me they know exactly how far they still have to go. Watching closely. This is the race that matters in 2026 - not which model scores highest on benchmarks, but which agent actually does useful work without breaking things. Like Reply 145 Reactions 146 Reactions Vincent Nwaka 1mo Report this comment Honestly, so much is coming from Claude right now that I have not even had the chance to try the last update yet. But that is actually a good problem to have. The idea of assigning a task from your phone and coming back to finished work on your computer is not just a productivity win. It fundamentally changes how you think about your day. You stop being the executor and start being the decision maker. For anyone working in Data and AI, the ability to automate routine tasks like pulling weekly reports or scanning emails means more headspace for the work that actually requires human judgement. 👏 Like Reply 8 Reactions 9 Reactions Anu Verma 1mo Report this comment Hot take, the interesting part is not that an agent can click around your desktop, it’s that this pushes software back toward the UI as the API. That looks flexible in demos, but in enterprise environments it gets fragile fast because permissions, audit trails, and tiny interface changes become model-breaking events. The AI part is almost the easy bit here, the real question is whether this creates a reliable work layer or just a very impressive workaround for badly integrated systems. Like Reply 4 Reactions 5 Reactions Tolani Omitokun 1mo Report this comment I love that I actually tried this while exploring before even seeing this post! The team is working overtime!! if you are curious about how to use all this new features, check my most recent post! No more previous content No more next content Like Reply 2 Reactions 3 Reactions Will Tygart 1mo Report this comment Running it right now actually — had Claude open my LinkedIn, search hashtags, read posts, and comment in my voice while I watch. The "assign from your phone and come back to finished work" is where this gets genuinely different. It's not assistance anymore, it's delegation. Like Reply 2 Reactions 3 Reactions David Malovecky 1mo Report this comment The phone-to-desktop workflow is the detail that matters most here. I go for walks when I need to think through a problem. Fresh air, no screen, just the problem. The ideas usually come halfway through. Until now, the gap between having the idea and acting on it meant sitting back down, context-switching, and losing momentum. Now I can send the task from the park and come back to work already in progress. That changes the rhythm of how I work with Claude Code entirely. Like Reply 1 Reaction 2 Reactions Des Raj C. 1mo Report this comment This is a massive jump for Dev productivity but the real enterprise hurdle isn't the capability, it's the observability. Shipping agentic workflows to production is one thing, but giving a model raw computer access in a compliance-heavy environment means we need a whole new layer of real-time audit logs that most security stacks just aren't built for yet. The moment an agent misinterprets a UI element in a live CRM, the recovery path is a nightmare compared to structured API calls. Like Reply 4 Reactions 5 Reactions Pedro Freitas 1mo Report this comment Claude, please give us a break ✊ With updates like this, I’m never going to stop improving my skills and agents 🤣 Like Reply 22 Reactions 23 Reactions PromptCraft AI Hub 1mo Report this comment This feels like the moment AI stops being something we use and starts becoming something that actually works for us. But the real shift isn’t the feature—it’s the mindset. Most people will test this and move on. A few will realize this changes how work itself is structured. Because once AI can execute across tools, the advantage is no longer access,it’s how well you design the system behind it. Same AI, completely different outcomes. The gap is about to widen fast. 👀 Like Reply 1 Reaction 2 Reactions See more comments To view or add a comment, sign in