Windsurf's Cascade is built for long, autonomous flows — which makes its forgetfulness sting more, not less. The agent that just refactored half your codebase can't remember why it made those choices tomorrow. Local memories help within a machine, but the moment you switch computers or start a fresh workspace, the accumulated understanding is gone, and Cascade re-derives your conventions from scratch. This guide shows what persistent memory for Windsurf looks like in practice — what to store, what it changes, and exactly how to wire it up. New to the idea? Start with the complete guide to persistent memory for AI.
Why Windsurf needs persistent memory
The problem isn't the model — it's statelessness. Windsurf is only as consistent as the context it's handed each session, and without a durable memory that context resets to zero every time. Persistent memory is a store outside the model that keeps the facts, decisions, and preferences you build up, and surfaces them by meaning when they're relevant again — so Windsurf stops re-learning your project and starts compounding what it knows.
What your Windsurf setup should remember
- The plan and the reasoning behind a multi-step refactor, so a follow-up session continues instead of restarting.
- Project conventions Cascade should apply everywhere: structure, error handling, naming.
- Decisions made mid-flow that live in no file yet — "we're dropping the legacy adapter after the migration."
- Operational facts: how the deploy pipeline behaves, which service owns which queue.
What it changes, concretely
Before: you finish a big Cascade session, come back the next day, and spend the first ten minutes re-establishing where you were. After: Cascade recalls the plan, the decisions, and the open threads — the second session starts three steps ahead of the first.
How to add persistent memory to Windsurf
- Add the Kit for AI MCP server in Windsurf's MCP configuration with your API key.
- Cascade gets remember and recall as tools and uses them autonomously during long tasks.
- Use one namespace per project so a big monorepo's context doesn't bleed into a side project.
- The same memory is readable from Cursor, Claude Code, or a script — it's cloud state, not a per-editor cache.
One memory across every tool
Because the memory lives in the cloud, it isn't locked to Windsurf. A decision you save here is recalled by every other MCP-capable tool and by your own scripts over the API — switch models or editors and your memory follows you. That's the difference between a per-app memory silo and a real memory layer for AI, and it's why the same account also handles document ingestion and cited search.
FAQ
Does Windsurf have persistent memory built in?
Windsurf keeps context within a session and through static config files, but it has no durable, cross-session memory of the decisions and preferences you build up — that resets every time. Persistent memory adds a store outside the tool so those facts survive restarts and follow you across machines.
How do I give Windsurf long-term memory?
Add the Kit for AI MCP server to Windsurf with an API key — remember and recall become tools it calls on its own. It takes a couple of minutes, and the memory is shared with every other MCP-capable tool you use.
Related guides
Go deeper: cloud persistent memory for AI agents · share one memory across AI models — or read the pillar guide to persistent memory for AI.
Add persistent memory to Windsurf in minutes — Start free, drop in the MCP server with your API key, and remember/recall become tools Windsurf calls on its own.