Store a long-term memory that persists across sessions AND across every AI tool the user has connected to Mnemoverse (Claude, ChatGPT, Cursor, VS Code) — write once, recall everywhere. Call this PROACTIVELY the moment the user states a preference, makes a decision, or you learn a durable fact (people, roles, project setup, a lesson). Don't wait to be asked. Do NOT use it for transient chatter, secrets, or anything only relevant to the current turn. Behavior: an importance gate may filter low-value writes, so the result tells you whether the memory was stored or filtered. Write content as a self-contained statement that still makes sense when recalled out of context.

Search your long-term memory before answering anything that may have come up before — user preferences, past decisions, project setup, people, or earlier context. This memory is shared: it persists across sessions and across every AI tool the user has connected (Claude, ChatGPT, Cursor, VS Code). ALWAYS check here first when you're unsure whether you already know something; no need to call it for general world knowledge you already hold. Returns matches ranked by relevance (semantic similarity plus learned concept associations); each result carries an id you can pass to memory_feedback or memory_delete.

Report whether memories returned by memory_read were actually helpful. This is a learning signal, not a log: positive feedback raises a memory's ranking so it surfaces faster next time (across all of the user's tools), negative feedback lets it fade. Call it right after you act on (or reject) recalled memories, passing the ids from the memory_read results.

Get an overview of the stored memory: total count, episodes vs consolidated prototypes, number of learned associations, the list of domains, and average quality scores. This memory is shared across all AI tools the user has connected to Mnemoverse. Use it to orient yourself, to confirm the exact domain name before a delete, or when the user asks what you remember. Read-only — changes nothing.

Permanently delete ONE memory by its atom_id — irreversible, the memory is gone for good. Use only when the user explicitly asks to forget something specific, or to correct a fact you stored wrongly. Get the atom_id from a memory_read result. To clear an entire topic at once, use memory_delete_domain instead. Never delete on your own initiative — require an explicit user request.

Permanently delete EVERY memory in one domain — irreversible, and far more destructive than memory_delete. Use only on an explicit user request to wipe a whole topic, e.g. 'forget everything about project X' or 'wipe my benchmark experiments'. First run memory_stats to confirm the exact domain name, then pass it together with confirm=true (a deliberate safety interlock). Never call this speculatively or to 'clean up' on your own — only when the user explicitly asks.