Workspace memory
What it means that a workspace remembers, who adds entries and who approves them, who sees an entry, and how to add, delete or switch off the daily reflection.
A workspace has a memory for things that last: preferences, ways of working, recurring objections, how someone likes to be addressed. Agents use these entries when they work, so you do not have to explain the same thing again. An entry is one sentence with its origin and a visibility. This is not a log of everything that happens.
Who adds entries?#
- The Curator, once a day. It looks back at recent changes, at approvals and what the advisor edited before approving them, at topics handed over from sparring and at journal entries the person has shared with the advisor. It proposes a few entries. It does not read private chats or the private journal; it sees them only where they were handed over or shared.
- You, in the chat. Say "remember that ..." and Arvemo stores the sentence. A sentence that already exists is not stored twice.
Do I have to approve entries?#
The advisor releases every proposal of the Curator. In the Memory card on the Brain page, each proposal shows its origin and a visibility choice. Accept makes it active, Reject discards it, and Delete removes an active entry later. Proposals appear dashed until released. What the advisor tells Arvemo to remember is active at once. What a client says, Arvemo files as a proposal for the advisor to release. The Brain page is the advisor's own; the client does not see it. It also exists on the Mac and iPhone.
Who sees an entry?#
| Visibility | Who sees it |
|---|---|
| Only the advisor | the advisor side only |
| Whole workspace | advisor and client side |
| One person (picked per entry) | that person and the advisor |
In the app the labels carry the names involved.
Internal fits politics, people's weaknesses and the advisor's own style. Workspace-wide fits neutral working agreements. The client side sees only released entries. A client can store only workspace-wide entries or entries about herself, and only as proposals. A person with the role Read cannot add any.
What does the client see about herself?#
On the web, the Mac and the iPhone, clients find What the workspace knows about you on the Workspace page. It lists the released entries that are workspace-wide or only about her, each with its origin. The page states that the workspace remembers only what she shared or what is visible in it, never her private journal or private conversations. She can delete entries about herself; a Responsible can also delete workspace-wide ones.
Switch the daily reflection off#
Open Settings, then Agents, and set Memory to off or daily. The setting applies to all workspaces of your organisation. It also stays off if the Curator itself is switched off. The same card on the Brain page has Reflect to start a reflection at once.
Workspace memory and organisation memory#
Workspace memory belongs to one workspace. Your organisation also has a memory of how its advisors work, shared across workspaces. It is for the advisor side only; clients never see it. A released style entry of yours can become part of it.
Common questions#
Does ARVEMO read my private chats?
Not on its own. The daily reflection does not read private chats or the private journal. It only uses what was handed over to the advisor (a sparring topic you passed on, a journal entry you shared) and what happens in the workspace. Something you ask Arvemo to remember is stored as an entry.
How do I tell Arvemo to remember something?
Write it in the chat, for example "remember that the board wants one page".
Can the client delete entries about herself?
Yes, on What the workspace knows about you.
Do entries last forever?
No. Proposals that nobody decides on lapse, and entries have a limited lifetime.