Memco Personal Memory

By Memco Labs inc.

Get New App Alerts

Open in ChatGPT

Open in Claude

Overview

Memco Personal Memory gives your AI assistants persistent memory about how you work. Preferences, habits, project context, and working style carry across Claude, ChatGPT, Cursor, and any other MCP-compatible client. Every agent starts where the last one left off. Your memories are private, hosted, and fully under your control. The Memco dashboard at memco.ai lets you view, edit, and delete individual memories at any time. There is no local setup, no file to manage, and no data locked inside a single tool. Memco Personal Memory works alongside Memco Shared Memory. Personal memory holds what is yours: how you like responses formatted, your editor setup, your role and working context. Shared memory holds what your team should know: validated solutions, operational decisions, institutional knowledge. Same platform, complementary scopes. Use one or both. Free to use. Sign up at memco.ai.

Tools

delete_memory

ChatGPT

Permanently delete one personal memory by its external_id. There is no undo. Call when: a saved preference is wrong or no longer true and no edit would fix it. Prefer update_memory when the note should change rather than disappear.

get_memory

ChatGPT

Fetch one personal memory by its external_id and return the note itself: its title and content, under a heading naming the external_id and the date it was last updated. Call when: start_session or a listing showed you an external_id whose note you now need in full — those listings carry only a one-line hook. Find the external_id with list_memories or search first; this tool does not search.

list_memories

ChatGPT

List your personal memories, newest first, as external_id: title. Pass page/page_size to paginate, keyword to grep insight text, or tags to filter. Call when: you want to see everything saved for this user, or to find the external_id of a note you intend to get, update or delete. Use search instead when you know what you are looking for but not what it was called.

search

ChatGPT

Search your personal memory for notes about this user. Set keyword_only to rank on exact terms instead of meaning, which suits identifiers and exact strings. Call when: you need to know how this user wants something done and did not already load it at start_session, or you are about to save a preference and want to check whether one already covers it. Use list_memories instead to browse everything.

start_session

ChatGPT

Load this user's identity and preferences: their type:user memories, shown as description + external_id, plus the tags in use with a count each. What it returns is your standing context for how to work with them. Call when: at the start of a session, before your first substantive reply. Follow up with get_memory for any preference you need in full, and search for anything it did not load.

update_memory

ChatGPT

Edit an existing personal memory in place by its external_id, like editing a file: replace a unique substring of the content (old_string → new_string, or replace_all), prepend/append text, and/or set a new title or description. Untouched fields and tags are preserved. Call when: a preference you already saved has changed or needs a detail added and the rest of the note should survive. Use write_memory with the same external_id when the whole note should be replaced.

write_memory

ChatGPT

Save a note to your personal memory — this user's own preferences and habits, which no one else can read. Call when: you learn a durable, private preference of this user's: how they like you to respond, their editor and workflow settings, or a personal fact that should never be shared. One preference per memory. Anything a teammate could reuse belongs in shared memory instead. Supply a stable external_id to name it — rewriting the same external_id replaces the note wholesale, so reach for update_memory when only part of it should change. Add a one-line description: it is the hook shown in your startup listing, and what semantic search matches on.