Skip to content

fix(computer-use): scale scroll amounts #23

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
juecd merged 3 commits into from
Jun 19, 2025
Merged

fix(computer-use): scale scroll amounts #23

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
3 commits
Select commit Hold shift + click to select a range
04fc291
fix(computer-use): scale scroll amounts in TypeScript
matthewjmarangoni Jun 11, 2025
2b13dec
fix(computer-use): python template scrolling scale
matthewjmarangoni Jun 14, 2025
2bb650f
fix(computer-use): print scroll distance in pixels in TypeScript
matthewjmarangoni Jun 14, 2025
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
20 changes: 15 additions & 5 deletions templates/python/computer-use/tools/computer.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -316,16 +316,26 @@ async def __call__(
await self.page.mouse.move(x, y)

# Map scroll directions to Playwright's wheel events
page_dimensions = await self.page.evaluate(
"() => Promise.resolve({ h: window.innerHeight, w: window.innerWidth })"
)
page_partitions = 25
scroll_factor = scroll_amount / page_partitions
page_width = page_dimensions['w']
page_height = page_dimensions['h']

delta_x = 0
delta_y = 0
if scroll_direction == "up":
delta_y = -scroll_amount * 100
delta_y = -scroll_factor * page_height
elif scroll_direction == "down":
delta_y = scroll_amount * 100
delta_y = scroll_factor * page_height
elif scroll_direction == "left":
delta_x = -scroll_amount * 100
delta_x = -scroll_factor * page_width
elif scroll_direction == "right":
delta_x = scroll_amount * 100
delta_x = scroll_factor * page_width

print(f"Scrolling {abs(delta_x) if delta_x != 0 else abs(delta_y):.02f} pixels {scroll_direction}")

await self.page.mouse.wheel(delta_x=delta_x, delta_y=delta_y)
return await self.screenshot()
Expand Down Expand Up @@ -385,4 +395,4 @@ async def __call__(

return await super().__call__(
action=action, text=text, coordinate=coordinate, key=key, **kwargs
)
)
10 changes: 9 additions & 1 deletion templates/typescript/computer-use/tools/computer.ts
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -190,11 +190,19 @@ export class ComputerTool implements BaseAnthropicTool {
await this.page.waitForTimeout(100);
}

const amount = scrollAmountValue || 100;
const pageDimensions = await this.page.evaluate(() => {
return { h: window.innerHeight, w: window.innerWidth };
});
const pagePartitions = 25;
const scrollFactor = (scrollAmountValue || 10) / pagePartitions;

if (scrollDirection === 'down' || scrollDirection === 'up') {
const amount = pageDimensions.h * scrollFactor;
console.log(`Scrolling ${amount.toFixed(2)} pixels ${scrollDirection}`);
await this.page.mouse.wheel(0, scrollDirection === 'down' ? amount : -amount);
} else {
const amount = pageDimensions.w * scrollFactor;
console.log(`Scrolling ${amount.toFixed(2)} pixels ${scrollDirection}`);
await this.page.mouse.wheel(scrollDirection === 'right' ? amount : -amount, 0);
}

Expand Down