Draw on screenshots right in the task
Open an image on a task and mark it up: circle the button that's in the wrong place, box the banner that should go, draw an arrow to where things belong. Everyone on the task sees your markings, and so do AI agents that read the screenshot.
Try it: draw on the screenshot, click a marking to resize it, then Save and close.
In the task's activity
Ida · just now
Two things on checkout, see the screenshot.
What an agent reads
checkout-qa.png (image/png, 412 KB) · with 4 markings the team drew on it
Draw right away
Open an image on a task and the pen is ready. No mode to switch on.
Lines, boxes, circles
Eight colours. Pick a colour with a marking selected to recolour it.
Click to adjust
Drag the handles to resize, drag the shape to move it, Backspace deletes it.
Eraser, undo, reset
The eraser removes any marking it passes over, saved ones too. ⌘Z undoes.
Save or discard
Save and close keeps the session; Discard changes goes back to what was saved.
Agents see it too
An agent that opens the screenshot gets it with your markings drawn in.
What's new
- Drawing is always on in the image preview: pen, line, rectangle, circle and eight colours, in a tool island next to Download
- Click a marking to select it: drag the handles to resize, drag it to move, Backspace deletes it
- The eraser removes any marking it passes over, including ones saved earlier; Reset clears them all and ⌘Z undoes
- Save and close keeps your markings, Discard changes goes back to what was saved, and leaving with unsaved markings asks first
- Markings show on the image's thumbnail in the activity and the files panel; the original file is never changed
- AI agents that open the screenshot get it with the markings drawn in, so they see exactly what QA pointed at
QA feedback on a screen is easiest to give on the screen itself. Until now that meant opening the screenshot in another app, drawing on it, saving a copy and uploading it again, so the task ended up with two versions of every image. Now you draw where the image already is. Open it on the task and the pen is ready: circle what's wrong, box what should go, draw an arrow to where it belongs. Switch to a line, a rectangle or a circle, pick one of eight colours, and click any marking later to resize, move, recolour or delete it.
Markings are saved next to the image, not burned into it. The original stays as it was, Download still gives you the untouched file, and the eraser can remove a marking someone drew last week as easily as one you drew a minute ago. A markup session starts when you open the image and ends when you close it: Save and close keeps it, Discard changes throws away what you did in this session, and Hourtick asks before you leave with markings you haven't saved. Anyone who can work on the task can mark up its images, the same as commenting.
Your markings reach the whole team, including its agents. The thumbnail in the activity and in the files panel shows them, and when an AI agent opens the screenshot through Hourtick's MCP server, it gets the image with the markings drawn in, and the answer says how many there are. So when you hand the fix to an agent, it sees the circle around the button, not just the bare screenshot.
There's more in this release. Images in the preview now fill the window, so Retina screenshots no longer show up tiny. A task has a search field in its top bar (or ⌘F) that finds and highlights text in the description, checklist, subtasks and activity. Comments on tasks render Markdown like chat does, with clickable task links, and an agent's @mention in a comment now notifies the person. Files you're about to send in chat or a comment can be previewed before you send them, and checkboxes have clearer borders. All of it is included on every plan.
More in Tasks
Track your first hour in a minute.
Free for your whole team, forever. Every feature included.