very slow going through the steps and the process ends with page screenshot timeout

[samer-alisaleh]

Hi,

Thanks for providing the code. Im trying to run it as is without any change but its been a struggle to get it to do anything right. Here are the steps I followed:

1- Download the repository as a zip package
2- Unzip
3- install the required packages
4- Open the code in VisualCode
5- I opened new terminal and typed the following command:

python cli.py --computer local-playwright
6- In the input prompt I typed the following task:

"go to cnn.com, find the main headline then summarize and print"

7- Chrome session was open to the Bing.com
8- The agent typed cnn.com in the search bar
9 - The page went to cnn.com - so far so good.
10- The agent seems to go off track where multiple browser sessions opened for cnn.com.
11- The loop seem to get stuck on different actions and repeat.
12- After few minutes of waiting and repeating, the code existed with the following error:

   raise rewrite_error(error, f"{parsed_st['apiName']}: {error}") from None
playwright._impl._errors.TimeoutError: Page.screenshot: Timeout 30000ms exceeded.
Call log:
  - taking page screenshot
  -   - waiting for fonts to load...
  -   - fonts loaded

Here is the full action taken by the agent as appeared in the terminal:

screenshot({})
click({'button': 'left', 'x': 451, 'y': 177})
keypress({'keys': ['CTRL', 'L']})
type({'text': 'cnn.com'})
keypress({'keys': ['ENTER']})
wait({})
click({'button': 'left', 'x': 459, 'y': 299})
wait({})
click({'button': 'left', 'x': 464, 'y': 300})
wait({})
click({'button': 'left', 'x': 356, 'y': 45})
keypress({'keys': ['CTRL', 'L']})
type({'text': 'cnn.com'})
keypress({'keys': ['CTRL', 'A']})
type({'text': 'cnn.com'})
keypress({'keys': ['CTRL', 'A']})
type({'text': 'cnn.com'})
click({'button': 'left', 'x': 372, 'y': 300})
wait({})
click({'button': 'left', 'x': 438, 'y': 38})
click({'button': 'left', 'x': 744, 'y': 48})
wait({})
click({'button': 'left', 'x': 379, 'y': 45})
click({'button': 'left', 'x': 743, 'y': 44})
keypress({'keys': ['CTRL', 'T']})
click({'button': 'left', 'x': 345, 'y': 48})
click({'button': 'left', 'x': 329, 'y': 49})
keypress({'keys': ['CTRL', 'A']})
type({'text': 'cnn.com'})
keypress({'keys': ['CTRL', 'A']})
click({'button': 'left', 'x': 403, 'y': 333})
wait({})
click({'button': 'wheel', 'x': 428, 'y': 298})
wait({})
click({'button': 'left', 'x': 374, 'y': 299})
wait({})
click({'button': 'wheel', 'x': 633, 'y': 297})
scroll({'x': 827, 'y': 211, 'scroll_x': 0, 'scroll_y': -367})
click({'button': 'left', 'x': 28, 'y': 56})
click({'button': 'left', 'x': 40, 'y': 73})
click({'button': 'wheel', 'x': 451, 'y': 297})

Total time between start to finish 10-15 minutes.

Not sure what I'm doing wrong. I appreciate your feedback and guidance.
Thanks

[katia-openai]

Sorry about this, it was probably linked to the issue fixed in #11 - the screenshot was actually returning content from the 1st page. Should work now!