Lovable: Memory That Remembers Your Preferences and Decisions
LovableLovable now remembers the preferences, decisions, and facts users state while building, and applies them in later conversations without reminders. Project memory is shared by everyone who can edit a project, while personal memory covers how a user wants Lovable to work in every project. Memory settings show a plain-language summary, let users correct items through chat, and let them choose which memory Lovable uses or disable it.
Key Takeaways
- Lovable now remembers the preferences, decisions, and facts users state while building, and applies them in later conversations without being reminded.
- Project memory is shared by everyone who can edit a project, so team decisions persist across collaborators.
- Personal memory follows the user into every project and captures how they want Lovable to work, such as reply length and technical depth.
- Memory is transparent, with a plain-language summary in workspace and project Memory settings rather than a hidden store.
- Corrections happen through chat: selecting text and choosing Change in chat lets users fix something Lovable remembered wrongly.
- Users stay in control, choosing which memory Lovable uses or disabling it altogether.
Lovable Now Remembers What You Tell It
Lovable introduced Memory, a feature that picks up the preferences, decisions, and facts a user states while building and applies them in later conversations. Users no longer need to repeat the same instructions at the start of each chat. Lovable carries the context forward on its own.
Two Kinds of Memory
Memory comes in two scopes, each with a different audience.
Project memory
Project memory belongs to a single project. Lovable adds to it as the conversation goes on, and it is shared by everyone who can edit that project. Decisions made by one collaborator therefore carry over to teammates working in the same project.
Personal memory
Personal memory covers how a user wants Lovable to work with them, and it applies in every project they build. Typical examples are how long replies should be, how technical the explanations can get, and whether Lovable should ask before making a small change. Other members of a project cannot read a user's personal memory.
Seeing and Controlling What Lovable Remembers
Workspace and project Memory settings show a plain-language summary of what Lovable has stored. To correct an item, a user selects the text and chooses Change in chat, which opens a conversation to fix it. The same settings let users choose which memory Lovable uses or disable memory entirely.
Why It Matters
Teams building with Lovable often lose time restating conventions, product decisions, and working style. By separating shared project knowledge from individual working preferences, Lovable keeps collaborators aligned on what is being built while letting each person keep their own style across projects.