Asking questions

How to ask

Ask the way you would ask a colleague, and mention memors so Claude checks the memory first:

What does memors say about how we handle failed payments?

Check memors: which version of the gRPC library do we use, and what changed in it?

Claude searches and reads the most relevant passages. It then answers and names its sources.

Starting a chat

memors-mcp offers a ready-made starter, ask, that keeps Claude to the memory for the rest of the conversation:

• Claude Code: type /mcp__memors__ask and press Enter. Claude says in a sentence what the memory holds and asks what you want to know; then ask your questions as usual. Put the question in a message of its own: a word after the command is read as a shelf name, so /mcp__memors__ask billing keeps the chat to the billing shelf.
• Claude Desktop: click +, choose memors, then ask. Leave the shelf empty to search everything, and type your question in the message.

The middle of the command is the name memors-mcp was connected under. A memory added as memors-client-b (see Keeping work separate) starts with /mcp__memors-client-b__ask, and one connected from the menu-bar app as memors-work with /mcp__memors-work__ask. In the list that appears when you type /, they show as memors-client-b:ask (MCP) and so on.

Claude then answers only from what the memory holds. Every statement ends with a footnote naming the saved item and its memo:// address. If the memory has nothing on a point, Claude says so instead of guessing, and it tells you when an answer rests only on something Claude saved itself. To narrow a question, say so in the chat, for example “only version 1.64”.

What comes back

Each result carries:

• Where it came from: the document title and, for web pages, the original address.
• Who saved it: you, Claude, or a checked source. See Trust.
• When it was saved, and whether a newer version exists.
• How strong the match is: strong, moderate or weak. A weak match is a hint, not an answer, and Claude should treat it that way.

Asking why

Every search can explain itself. Ask:

Why did that come up first?

Claude can show which kinds of matching found each result. Some results match your exact words, some match a name or code identifier exactly, and some match the meaning even when the words differ. Claude can also show how each result scored. The same explanation is visible in the browser view. See Looking inside.

When nothing is found

If the memory holds nothing relevant, memors-mcp says so plainly and does not offer the closest guess. That is deliberate. “We don’t have this” is more useful than a confident wrong answer. Claude then knows to look elsewhere. Once it finds the answer, ask it to save it, so the next person does not have to search again.

Too many results

If a question matches a lot, the answer says that some results were left out and suggests how to narrow it, for example by version or by shelf. You can say:

Only look at version 1.64.

Only search the billing shelf.

Asking about the past

Because old versions are kept, you can ask what the memory said at an earlier date:

What did memors say our database version was on 1 March?