Answers
Arroway vs Supermemory: what is the difference?
Supermemory describes itself as context infrastructure for AI agents that gives an agent long-term memory of every user. According to its documentation, you send it conversations, documents, files or URLs; a memory pipeline extracts facts, builds a profile and a graph, and search returns memories, chunks or both. It offers a local self-hosted setup, has an MCP server and integration pages for Claude Code and Cursor, and isolates data by namespace, which its documentation calls a hard boundary for a user, tenant or project. Arroway is not built from what was said to an agent: 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 decides what is current. In Supermemory, the pages I read say its learning model decides what is important and when to forget. 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
Remembering what was said, or recording what was decided
Supermemory's documentation describes memory as something it derives: content goes in, and a pipeline extracts facts, a profile with static and dynamic parts, and a graph. What it holds is built from what a user or an application fed it. Arroway starts from a different object: not what was said, 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 decides what is current
In the Supermemory pages I read, facts change over time: the documentation's profile example shows a fact moving from one employer to another, and it describes endpoints to remove facts by id or by semantic match, with a dry-run preview. It says the learning model decides when to forget, and it does not describe a person approving what stands. Its isolation is by namespace, a boundary for a user, a tenant or a project, and the pages I read do not describe shared team memory with approvers. In Arroway the project is the shared object: the people on it and the assistants they use read the same rules, each one carrying who approved it and the condition that ends it.
When to use Arroway
Use Arroway when the question is not what the assistant remembers about someone, but what is in force for the team: who decided it, what ends it and what it replaced. A newer decision names the rule it retires, a person approves it, and the old rule leaves the next read while staying in the history. Nothing is built: 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. The first step is the install page.
What it looks like in practice
A five-person software company ships a customer assistant that uses Supermemory to remember each customer's conversations. Internally, the same five people use Claude, Codex and Cursor to build the product. On Monday the team decides that new accounts get a 14-day trial. On Thursday the founder cuts it to 7. What the customer assistant remembers is about each customer and what they said. The trial length is a decision of the company: the assistants the team uses to write the pricing page, the emails and the code each read whatever someone gave them, and nobody can tell which number the team stands behind. In Arroway, Thursday's rule is recorded as replacing Monday's, with the founder's name on it. Claude, Codex and Cursor open their next session with the 7-day rule, the customer assistant's builder can read it over MCP, and the history shows the old figure, when it changed and who changed it. We have not built anything with Supermemory; what is said of it here comes from its documentation.
Questions people ask about this
- Does Arroway replace Supermemory in my product?
- No. Remembering what each user said 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 product should act on them, it reads the project over MCP like any other tool.
- Can Arroway hold a personal memory of me?
- Arroway serves people and agents, and what it records is what a project decided, not a profile of one person. An agent working alone is the product, not a reduced version of 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
Supermemory documentation (supermemory.ai/docs), the introduction and the How Supermemory works page, read on 9 October 2026 — Supermemory is described only as that documentation describes 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 Mem0 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 company in the example is illustrative.