Use of OS (ie my) installed Chrome browser as managed browser using CDP

[helicalchris]

I have danced around this and every variant I can find within crawl4ai to connect to my Chrome, not a Chromium it starts. I know playwright can do it, and its part of mainstream CDP use, but I cannot for the life of life of me work out how in crawl4ai to point at an existing Chrome instance and use that. I specifically want this because the site I am crawling is very very sensitive to bots and I need to have regular interaction with it, not just a recorded profile. Otherwise as soon as I do deep-crawling I get booted off it.

I can set CDP endpoints using crwl or in code, but the only time I got it to even faintly work is when I used crwl, which killed my Chrome instance and started its own Chromium! Otherwise it seems to say its working but I see no interaction with Chrome so I think that actually its starting a headless chromium to scrape and returning that result and totally ignoring the CDP URLs I provide (which is similar to the behaviour I see with crwl)

Is it possible? If not, could it be, as I can see in the code that connect_with_cdp is being used, and its all playwright based, so its very much feasible to I believe...

Chris