However, using this eats up all the context. I tried creating an agent for this using model sonnet[1m] for it. This kind of works but agents ends up doing a lot more than just using Chrome. It forgets that it is supposed to complete the flow or view console logs, take screenshots and give complied result back but it rather starts working on the issue in hand. Maybe it is because of poorly written Agent or poorly written Claude.md.
How is everyone working with playwright-mcp/Claude Chrome?
reify•1mo ago
It forgets! It doesn't forget, it doesn't have a memory or a human brain.
maybe!
Consider your own life. look back at your childhood and remember, running home from school on a beautiful sunny day, playing with your friends, remember the freedom. How many years ago was that? 20, 30, 40, 50, and you still remember it like it was yesterday
AI DOES NOT HAVE ANY MEMORIES SO IT CANNOT FORGET.
jaigupta•1mo ago
I am not sure if I need to write better CLAUDE.md (main context) or better agent file .claude/agents/chrome-tester.md for this case.
I get what you were trying to say but AI does have memory, fixed memory files that we can modify and we may also write creative prompts for dynamic memory for e.g. while troubleshooting keep adding learnings to learning.md file so that you know what all was tried. It also have in-built memory in the sense that it has a lot of built in knowledge. Okay, I admit to be incorrect on purely technical point of view but in just simpler terms.