Answers
Arroway vs MemoryLake: what is the difference?
MemoryLake describes a Memory Passport: one memory kept across ChatGPT, Claude, Gemini and any MCP-compatible agent. According to its own blog post of 18 January 2026, memories have six types, each entry records its source platform, source conversation and a confidence score, every create, update and delete is stored as an event in an ordered log, and a three-level conflict detection applies resolution rules when two platforms change the same memory. It also describes a shared space for teams, available cloud-hosted or on-premise. Arroway is not a portable copy of what you and your agents said: it is a shared memory of what the people on a project decided, which every AI tool you already use — Claude, ChatGPT, Codex, Cursor or anything over MCP — reads before it acts. The difference is who settles a conflict. In the MemoryLake page I read, conflicting team input is flagged for resolution and the page does not say who resolves it. In Arroway, assistants write proposals, a person approves what becomes a rule, each rule carries the condition that ends it, and a replacement names the rule it retires. Use Arroway when a decision has to reach several tools and people as the current one, with its owner. The install page is the first step.
Last updated October 9, 2026
A memory that follows you, or a decision that follows the project
The MemoryLake post I read describes a portable format, an adapter for each platform and a sync protocol, so that a memory made in one assistant is available in another. What moves is what you and your agents produced in conversation. Arroway starts from a different object: not what was said on each platform, but what the people on the project decided. An assistant that writes into it writes a proposal; it becomes a rule only when a person approves it.
Who settles a conflict
The post says conflicts are detected at three levels and that resolution rules apply when two platforms modify the same memory; it does not say who defines those rules. For team memory it says conflicting input is flagged for resolution, without saying who resolves it, and it does not specify write permissions for the shared space. It records provenance, a source and a confidence score, which says where an entry came from and how sure the system is, not who has the authority to decide. In Arroway that authority is part of the rule: a person approves it, the replacement names the one it retires, and the old rule leaves the next read while staying in the history.
When to use Arroway
Use Arroway when what has to move between tools is not your memory but the team's decisions: who decided it, what ends it and what it replaced. Nothing is built or synced: you install it in each tool you already use, sign in, and the next session opens by reading the project and closes by recording what was decided. Every assistant reads the same current version, so a conflict reaches a person as a proposal to approve instead of an entry to reconcile. The first step is the install page.
What it looks like in practice
A three-person consultancy uses ChatGPT, Claude and Cursor. On Monday one partner tells her assistant that every proposal above a given value needs a second partner's sign-off. On Thursday another partner tells his assistant that the threshold is higher for repeat clients. With a memory built from each person's conversations, the two statements sit side by side, and what the next session reads depends on how the conflict is resolved and by whom. In Arroway, Thursday's statement is a proposal until the partner who owns pricing approves it as replacing Monday's rule, with her name and the condition that ends it. ChatGPT, Claude and Cursor open their next session with the current threshold, and the history shows the old one, when it changed and who changed it. We have not used MemoryLake; what is said of it here comes from its published pages.
Questions people ask about this
- Does Arroway replace MemoryLake?
- No. Carrying what you and your agents said from one platform to another is a different job from holding the team's decisions. Arroway holds the decisions: who approved them, what is in force and what ended. If your agent should act on them, it reads the project over MCP like any other tool.
- Why not just sync every assistant's memory with the others?
- Syncing carries what each assistant heard, including statements that disagree. A decision needs someone to say which one stands. In Arroway a person approves it, with the condition that ends it.
- How do I try Arroway?
- Install it in the tool you already use — the install page has one route for Claude, ChatGPT, Codex, Cursor or any MCP tool — and sign in when the tool asks. From then on, sessions open by reading the project and close by recording what was decided.
Where this is verifiable
MemoryLake, "One Memory, Every AI: How Memory Passport Works" (memorylake.ai/en/blogs/cross-platform-memory-passport, 18 January 2026), and its page on the memory API, read on 9 October 2026. I did not find its developer documentation, so MemoryLake is described only as those pages describe it on that date, and its features can change in any release. Product documentation on this site (How it works, Install), the answers on Arroway vs Zep and on the best shared memory for AI agents and teams, and the sanction, supersession, expiry-condition and opening-read rules in the sanctioned product spec. Everything described here about Arroway is behaviour the tools apply today, not roadmap. The consultancy in the example is illustrative.