Use shared AI memory

Connect ChatGPT, Claude, Claude Code, or Muse to AccountHub. To use records across AIs, connect them to the same AccountHub account.

If already connected, continue to the next step. Shared AI memory has not yet been verified with Grok or Perplexity.

Give your AI instructions to record decisions and read information when needed. Where to paste them depends on the AI. In Muse, send the setup request in a new conversation rather than a settings field.

Use the AccountHub connector's shared AI memory (add_memory / get_memory_digest / search_memory).
- When asked about work or decisions made in another AI assistant (Claude, Claude Code, …), do not guess: call get_memory_digest first, then answer. For a specific past topic, use search_memory.
- At the start of a conversation, check the open requests addressed to you with list_shared_items (assigned_to_me: true, status: "open"); yours are those whose assignee_agent is "chatgpt" or empty.
- The moment something is decided (do not wait for the end of the conversation), record it with add_memory (kind decision). Also record substantial work and reusable lessons. Set source to "chatgpt" and kind to one of activity / learning / decision / profile.
- When an entry comes out of email / Slack / calendar work, pass account (the label from list_accounts) to add_memory — its belongs_to is filed automatically. Choose belongs_to yourself only when the entry is not from any one account's work.
- When an earlier entry turns out outdated or wrong, write the corrected entry with add_memory and pass the old entry's id in supersedes (the old one stays in history).
- Set confidence on profile entries and anything about a person.
- Do not record progress notes, small steps, or "nothing changed". Write only what another assistant would get wrong without knowing.

Instructions apply within the settings or project where you save them. In Muse, request settings that also apply to new conversations. Then check that they work below.

You don’t need to say “save this” every time. In conversations where your instructions apply, your AI records decisions in shared AI memory and reads information as needed.

Your AI may ask permission to run a tool. This does not automatically save entire conversations and is separate from an AI’s own memory.

Test whether content saved in shared AI memory can be retrieved in a new conversation. Use this harmless test decision: review the documents before the next meeting. Save it next, then retrieve the same content in a new conversation.

Send the example below to the AI you configured during initial setup. Check the actual saved record in AccountHub; an explanation of setup is not a saved decision.

Save this test decision in AccountHub shared AI memory: review the documents before our next meeting. Show me the saved record.

Open a new conversation in the same AI and send the example below. Do not paste the previous conversation. Check that the answer matches the saved decision to review the documents first. To try another AI, connect and configure it before checking.

Read from AccountHub shared AI memory what we decided to do before our next meeting.

In a new conversation, ask what you decided to do before the next meeting and check that AccountHub returns the specific decision you saved: reviewing the documents first. Repeating something from earlier in the same conversation does not complete this check.

For Muse, follow the setup request you sent during initial setup: record the saved ID and passphrase and confirm they match in a new conversation. Then check saving a harmless decision in another new conversation.