Comment on Learnings from building AI agentsparentComments−asdev1ythe code reviews can't be effective because the LLM does not have the tribal knowledge and product context of the change. it's just reading the code at face value−theonething1yIsn't it possible to feed that knowledge and context to it? Have it scan your product website and docs, code documentation, git history, etc?
Comments
the code reviews can't be effective because the LLM does not have the tribal knowledge and product context of the change. it's just reading the code at face value
Isn't it possible to feed that knowledge and context to it? Have it scan your product website and docs, code documentation, git history, etc?