Project memory is saved context an AI tool brings into later conversations in the same project. The model never retrains on your corrections; the tool saves a note, preference, or summary and brings it back as context in a new session. Correct a company name’s spelling or state a project principle once, and later sessions use it.
An assistant’s notebook for one client works the same way. After a few meetings, the assistant writes down what keeps coming up: the client hates long emails, the owner signs off on pricing. They keep notes so they don’t ask the same questions again.
Why you care
Project memory differs from project knowledge, the files attached to the project, and it’s smaller than a second brain, the larger place where notes and decisions compound. Memory is scoped to the tool and shouldn’t be the only source of truth. Keep it clean: correct wrong memories, update changed preferences, remove stale ones. Bad memory is worse than none, because it feels helpful while quietly steering the work wrong. Repeated work should get easier instead of restarting from zero every morning.