Contents:
- What It Actually Does
- How It Works, Step by Step
- Setting It Up on Your Mac
- The Annotation Tools Inside
- Writing Better Prompts With It
- Who Gets the Most From It
- How It Fits Claude, Cursor and ChatGPT
- Privacy, Security and Sensible Limits
- Honest Limitations
- Alternatives and When to Choose Them
- Should You Try It?
- Frequently Asked Questions
What the Lasso Mac App Actually Does
The Lasso Mac app is a free macOS tool that solves one small but constant frustration in modern AI work: telling an AI agent exactly which part of your screen you mean. Instead of taking a screenshot, opening an image editor, finding the pen tool, drawing a circle, exporting the file and then uploading it, you draw directly on your live screen and capture the annotated view in one motion. The Lasso Mac app turns a fiddly five-step chore into something that takes seconds, and that speed is the whole point.
The idea is simple, and that simplicity is why it matters. AI agents such as Claude, Cursor and ChatGPT can already see images, but they cannot see what you are thinking. When you paste a plain screenshot into a prompt and write “fix the button in the top right,” the agent has to guess which button you mean, especially on a crowded dashboard or a dense block of code. A red circle, an arrow or a numbered step removes that guesswork. The Lasso gives you that visual pointer without ever leaving the window you are working in.
Developer Nika released the Lasso Mac app on Hacker News on October 21, 2026, positioning it squarely as a tool for visual prompting. Visual prompting is the practice of showing an AI model what you mean rather than describing it in words alone, and it has become a daily habit for developers, designers and product teams who work with coding agents. The Lasso is built for exactly that moment: you are staring at an error, a layout bug or a confusing interface, and you want the agent to look at the same thing you are looking at.
One important clarification before we go further. The name Lasso is used by more than one product. There is a well-known Lasso window manager on Setapp, and there are marketing tools with similar names. The Lasso Mac app covered in this guide is none of those. It is Nika’s free screen-annotation tool for AI agent prompts, version 1.2.1 at the time of writing, and it does one job: annotate your screen and hand the marked-up capture to an AI agent.

How the Lasso Mac App Works, Step by Step
Using the Lasso feels less like using an editor and more like pointing at something in real life. The app lives in your Mac menu bar rather than cluttering the Dock, and it places a transparent drawing layer over whatever is on your screen. You activate it, mark up the screen, capture the result and drag it into your agent. There is no export dialog, no file picker and no separate editing window to manage.
How the Lasso Mac App Capture Shortcut Works
The core workflow of the Lasso runs like this. First, you bring up the drawing overlay on top of your current app, whether that is a code editor, a browser, a design tool or a terminal. Next, you draw with the annotation tools: circle the error message, highlight the misaligned element, draw an arrow to the button that behaves strangely, or number a sequence of steps in the order the agent should follow them. Then you press S to capture the annotated view. Finally, you drag that capture straight into the prompt box of Claude, Cursor or ChatGPT and add a short line of text explaining what you want.
That drag-and-drop finish matters more than it sounds. Many annotation workflows break at the handoff: the marked-up image ends up buried in a downloads folder, or you have to remember to attach it after writing your prompt. The Lasso keeps the image in your hand, figuratively speaking, from the moment you mark it up to the moment it lands in the prompt. The result is a prompt that contains both your words and your pointing finger, which is a far richer instruction than either one alone.
Another thoughtful detail is that the Lasso Mac app toolbar itself is excluded from the final screenshot. Anyone who has annotated a screenshot before knows the embarrassment of sending an image that includes the annotation app’s own buttons and menus. Here, the exported image contains only your screen and your annotations, which keeps the agent’s attention on the content rather than on stray interface chrome.
Setting Up the Lasso Mac App on Your Mac
Getting started with the Lasso is deliberately low-friction. It is a native, notarized Mac app, which means Apple has checked it for known malware, and it requires no account and no sign-up. You do not hand over an email address, you do not create a password, and there is no subscription screen standing between you and the tool. For a utility that will sit on top of your screen all day, that lack of account friction is a genuine advantage.
The Lasso Mac app requires macOS 14 Sonoma or later, and it supports both Apple silicon and Intel Macs. That coverage matters because plenty of working developers still use Intel machines, and plenty of free indie tools quietly drop Intel support. If your Mac runs Sonoma or a newer version of macOS, you should be able to install and run the Lasso without upgrading your hardware.
Permissions the Lasso Mac App Needs, and Why
Permissions are where many screen tools lose people’s trust, so it is worth being precise about what the Lasso asks for. Drawing on the overlay requires no special permissions at all: the app is simply painting on a transparent layer above your apps. Capturing a screenshot is different, because macOS guards pixel data behind the Screen Recording permission. The Lasso needs that permission to read the screen and save the annotated capture, and macOS will ask you to grant it the first time you capture.
That split is a good privacy design. It means the Lasso Mac app is not quietly recording your screen in the background; it only reads pixels when you deliberately press the capture key. According to the launch coverage, the app also carries out no in-app tracking. Combined with the no-account setup, the privacy story is about as clean as a screen tool can offer, though you should still think before annotating screens that show passwords, private messages or customer data, since anything you drag into an AI prompt leaves your machine.
The Annotation Tools Inside the Lasso Mac App
The Lasso ships with a small, focused set of annotation tools rather than a bloated editor. Each one maps to a specific way of pointing at something, and learning when to use each one is what turns a messy screenshot into a clear instruction.
- Freehand circles: the workhorse tool. Circle an error message, a broken layout element or a suspicious line of code. Circles are fast, unambiguous and easy for a vision model to spot.
- Highlights: use these when you want to draw attention to a region rather than a single point, such as a whole paragraph, a form section or a block of terminal output.
- Arrows: arrows show direction and relationship. Point from a button to the panel it should open, or from a value to the place it should appear.
- Numbered steps: number a sequence when order matters. “Step 1, step 2, step 3” tells the agent to fix things in a particular sequence instead of guessing your priority.
- Laser pointer: useful for live walkthroughs and screen shares, where you want to guide attention temporarily without leaving a permanent mark on the capture.
The discipline here is restraint. A screenshot covered in ten circles, four arrows and three highlights is harder for an agent to read than a clean screenshot with one circle. The best habit with the Lasso Mac app is to mark the single most important thing first, add a second mark only if the relationship between two elements matters, and let your short text prompt do the rest of the explaining. Think of the annotation as the subject line of your prompt and the text as the body.

Writing Better Prompts With the Lasso Mac App
The Lasso does not write your prompt for you, and that is fine, because the combination of a marked-up image and a short, specific sentence is remarkably powerful. The annotation answers the question “where?”, and your text answers the questions “what is wrong?” and “what should change?” Agents perform best when both questions are answered clearly, and worst when either one is vague.
Consider a typical debugging exchange. Without annotation, you might paste a screenshot and write: “This page looks wrong, can you fix it?” The agent has to scan the whole page, guess what you consider wrong, and may confidently fix something you never cared about. With the Lasso, you circle the misaligned card, draw an arrow to the gap beneath it, and write: “This card should sit flush with the card beside it, with no gap underneath.” The agent now has a target, a direction and a success criterion, all in one message.
The same pattern works for code review. Circle a function, highlight the line where the value goes wrong, and write: “This returns null when the list is empty; make it return an empty list instead.” For design feedback, circle the button, highlight the surrounding spacing, and write: “Make this primary action stand out from the secondary links.” In each case the Lasso annotation removes the most common failure in AI-assisted work, which is not that the model is incapable, but that it was aimed at the wrong thing.
Prompt Examples That Work Well With the Lasso Mac App
Here are a few reusable prompt patterns that pair naturally with the Lasso Mac app. Each one keeps the text short because the image already carries the location information.
- Debugging: “The circled error appears when I click the highlighted button. Find the cause and fix it, without changing the surrounding behaviour.”
- Layout: “The arrowed element overlaps the highlighted text on narrow screens. Fix the overlap and keep the desktop layout unchanged.”
- Code review: “Review only the circled function. Flag logic errors and suggest a simpler version.”
- Onboarding: “Explain what happens, in order, at each numbered step in this screenshot.”
- Comparison: “The circled section differs from the design I described. List every difference you can see.”
Notice that none of these prompts waste words describing where things are. That is the efficiency gain the Lasso buys you: your words go toward intent, constraints and success criteria, while the image handles geography. Over a full working day, that saving adds up to dozens of shorter, clearer prompts.
Who Gets the Most From the Lasso Mac App
The Lasso Mac app is most valuable for people who already talk to AI agents about things on their screen. Developers are the obvious first audience: debugging, code review and explaining a stack trace are all visual tasks, and a circled line of code is worth a paragraph of description. Designers and front-end engineers come next, because layout bugs, spacing issues and visual regressions are almost impossible to describe precisely in words alone.
Product managers and QA testers benefit too. Filing a bug report with an annotated screenshot and a one-line description is dramatically more useful than a ticket that says “the checkout page is broken.” Support teams can use the Lasso the same way: circle the exact field a customer is confused by, and the agent can draft a precise answer instead of a generic one. Teachers and technical writers can number the steps in a workflow screenshot and ask an agent to turn it into documentation.
There is also a quiet accessibility benefit. Not everyone finds it easy to describe a visual problem in precise technical language, whether because they are new to a tool, working in a second language, or simply tired at the end of a long day. Pointing at the problem with the Lasso lowers the language barrier: you show the agent what you mean, and you only need enough words to say what should change.

How the Lasso Mac App Fits Claude, Cursor and ChatGPT
The Lasso is deliberately agent-agnostic. It does not plug into one company’s product; it produces an annotated image that you can drop into any agent that accepts image input. The launch coverage names Claude, Cursor and ChatGPT as the headline examples, and that trio covers the three main ways people currently work with agents: a chat assistant, a coding agent inside an editor, and a general-purpose chatbot.
With Claude, annotated screenshots slot into the normal chat flow: attach the capture, add your instruction, and Claude’s vision capabilities read the marks alongside your text. With Cursor, the fit is even tighter for developers, because you are often already inside the editor when the problem appears; circling the offending code and dropping it into Cursor’s agent chat keeps you in the same context. With ChatGPT, the Lasso Mac app is handy for explaining interfaces, documents and web pages that are easier to show than to describe.
It is worth noting what the Lasso is not. It is not an agent itself, it does not run code, and it does not automate anything. It belongs to the growing layer of small tools that make humans better at directing agents, alongside session tools, memory tools and orchestration tools. If you follow the AI agent space, you will recognise the pattern: the agents keep getting more capable, and the bottleneck shifts to how clearly humans can point them at the right problem. The Lasso Mac app attacks exactly that bottleneck.
Privacy, Security and Sensible Limits
The privacy profile of the Lasso is genuinely reassuring for a screen tool, but it is not a reason to stop thinking. The app needs Screen Recording permission to capture, it requires no account, and its launch coverage reports no in-app tracking. That is a strong starting position, and it compares favourably with screenshot tools that demand an account before you can draw a single circle.
Still, the moment you drag an annotated screenshot into an AI agent, that image is processed by the agent provider under that provider’s own terms. If your screenshot shows a production database password, a customer’s personal data, an unreleased design or an internal dashboard, you have shared that content with an external service. The Lasso cannot protect you from that decision, and no annotation tool can. Before capturing, hide or blur sensitive panes, close unrelated tabs, and ask yourself whether the agent genuinely needs the full screen or just the circled region.
There is also a practical security habit worth building: grant Screen Recording permission only to tools you actually use, and review the permission list in macOS System Settings from time to time. The Lasso earns its permission by doing one visible thing with it, capturing when you press S, but that standard should apply to every screen tool on your machine, not just this one.
Honest Limitations of the Lasso Mac App
Because the Lasso Mac app is a small indie release, honesty about its limits matters more than hype. First, it is Mac-only. There is no Windows or Linux version mentioned in the launch coverage, so if your team works across platforms, this tool will not cover everyone. Second, it is young: it debuted on Hacker News on October 21, 2026, and the launch coverage itself notes that the Hacker News thread had minimal engagement at the time of writing. That is not a criticism of the tool, but it does mean there are no published download figures, no large user community and no long track record to lean on yet.
Third, annotation quality still depends on the agent reading the image. Vision models are good at spotting a bold red circle, but they can still misread a cluttered screenshot, confuse two similar elements, or miss a subtle highlight. The Lasso improves the odds, it does not guarantee them. If an agent misreads your annotation, simplify the screenshot, use a bigger circle, or crop to the relevant region before annotating.
Fourth, the Lasso Mac app does not edit, redact or blur anything for you. If you need to hide sensitive information, you must do that before capturing, either by closing the relevant window or using a separate redaction step. Finally, as with any free indie tool, the long-term roadmap is unproven: there is no announced pricing, no paid tier and no guarantee about future updates. That is a fair trade for a free, no-account utility, but teams that need support guarantees should weigh it accordingly.

Lasso Mac App Alternatives and When to Choose Them
The Lasso is not the only way to annotate a screenshot, so it helps to know when a different tool fits better. macOS itself offers a solid built-in path: take a screenshot with the system shortcut, open it in Preview, and use Preview’s markup tools to draw shapes and arrows. That route costs nothing and needs no new app, but it restores the multi-step workflow the Lasso removes, and it is slower when you annotate screenshots many times a day.
Dedicated screenshot tools such as CleanShot X and Shottr offer annotation plus scrolling capture, cloud sharing and redaction, and they suit people who need a full screenshot suite rather than a single-purpose overlay. Browser extensions and design tools offer annotation inside specific apps, which is handy when you only ever annotate web pages or design files. What the Lasso offers against all of these is focus: it does one thing, it does it from the menu bar, and it hands the result directly to an AI agent.
The right choice depends on frequency. If you annotate a screenshot once a week, the built-in macOS tools are fine. If you annotate screenshots ten times a day while working with coding agents, the seconds saved by the Lasso Mac app compound quickly, and the drag-into-prompt workflow becomes the deciding feature. If you need redaction, scrolling capture or team sharing, a fuller screenshot suite is the better pick, and you can still keep the Lasso for quick agent prompts.
Should You Try the Lasso Mac App?
The Lasso Mac app earns its recommendation on a simple test: does it remove a real, repeated friction, and does it do so without adding new problems? On both counts, it passes. The friction is real, because anyone who works with AI agents has wasted time describing a screen in words when pointing would have done. The solution adds no account, no subscription and no tracking, and it asks for only the permission it genuinely needs to capture.
Try the Lasso if you use Claude, Cursor or ChatGPT with screenshots more than a few times a week, if you debug interfaces or code visually, or if you file bug reports and design feedback regularly. Keep your expectations honest: it is a young, Mac-only, single-purpose utility, not a platform, and it will not fix an agent that misunderstands your underlying request. But as a pointing tool for the age of AI agents, it is one of the cleanest examples of a small tool solving exactly one problem well.
Our advice is to install the Lasso, grant it Screen Recording permission, and use it for a week on real work: debugging, code review, layout feedback and bug reports. If you find yourself reaching for it without thinking, it has earned its place in your menu bar. If you only annotate occasionally, the built-in macOS tools remain a perfectly good fallback, and you will have lost nothing by trying.

Frequently Asked Questions About the Lasso Mac App
Is the Lasso Mac App really free?
Yes. The Lasso launched as a free download with no account required and no subscription mentioned in its launch coverage. Because it is an indie tool, it is sensible to check the download page for any future pricing changes, but at the time of writing there is no paid tier standing between you and the annotation tools.
Which Macs can run the Lasso Mac App?
The Lasso Mac app requires macOS 14 Sonoma or later and supports both Apple silicon and Intel Macs. If your Mac runs Sonoma or a newer macOS version, it should be compatible, regardless of which chip generation you own.
Does the Lasso Mac App work with ChatGPT as well as Claude and Cursor?
Yes. The Lasso produces a standard annotated image that you drag into any agent prompt that accepts images, including Claude, Cursor and ChatGPT. It is not tied to a single provider, so you can use the same workflow across all three.
Is the Lasso Mac App the same as the Lasso window manager?
No. The Lasso Mac app in this guide is Nika’s free screen-annotation tool for AI agent prompts. The Lasso window manager sold through Setapp is a different product that arranges application windows, and it has nothing to do with annotating screenshots for AI agents.
What permission does the Lasso Mac App need?
Drawing on the overlay needs no special permission, but capturing a screenshot requires macOS Screen Recording permission, because the Lasso must read pixel data to save the annotated capture. Grant it in System Settings when macOS prompts you, and review screen-recording permissions periodically as good practice.
For more on the world this tool lives in, see our guide to why major websites are blocking AI agents, our explainer on Google’s Data Agent Kit, and our look at Claude’s dynamic workflows for running agent teams. To verify the launch details in this article, see the ClawdBytes report on the Lasso launch.
