mem0-remember

Stores user-provided facts into mem0 memory with type classification and metadata.

Updated Jun 29, 2026
One-click install
npx skills add https://github.com/nhatnguyen1122/Agent-Memory-Eval --skill mem0-remember-nhatnguyen1122
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: mem0-remember
Source: https://github.com/nhatnguyen1122/Agent-Memory-Eval/tree/main/mem0/integrations/mem0-plugin/.opencode-plugin/opencode-skills/mem0-remember
Command: npx skills add https://github.com/nhatnguyen1122/Agent-Memory-Eval --skill mem0-remember-nhatnguyen1122

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? Capturing explicit decisions, preferences, and learnings during a coding session is easy to forget, and valuable context gets lost between sessions. This Skill stores user-stated facts verbatim into mem0 long-term memory with proper classification so they can be recalled later. ## Core Features & Use Cases - Verbatim Memory Storage: Saves the exact user-provided text via the mem0 add_memory API with infer disabled, since no extraction is needed. - Automatic Type Classification: Classifies content as decision, anti_pattern, user_preference, convention, task_learning, or environmental based on content signals. - Async Write Confirmation: Polls get_event_status with the returned event_id to confirm the write succeeded and report the memory ID. - Use Case: During a session the user says "remember that we always use pnpm instead of npm in this repo" — the Skill classifies it as a convention, stores it with branch and confidence metadata, and confirms the saved memory ID. ## Quick Start Ask the assistant to remember a fact, for example: remember that we decided to deploy with Docker Compose on staging.

Frequently Asked Questions about mem0-remember

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I save a memory in mem0 from a chat command?▼

Provide the text after the remember command, and the Skill calls add_memory with your user_id, app_id, and metadata. The text is stored verbatim with infer set to false because you stated the fact explicitly.

How does mem0 memory type classification work?▼

The content is matched against signal phrases: decisions and always-use statements become decision, failure warnings become anti_pattern, preferences become user_preference, conventions become convention, and setup or config content becomes environmental. Anything unmatched defaults to task_learning.

Why does add_memory return an event_id instead of a memory_id?▼

mem0 writes are asynchronous, so add_memory returns an event_id for the queued operation. You must call get_event_status with that event_id; when the status is SUCCEEDED, the result contains the actual memory ID.

What metadata is stored with each mem0 memory?▼

Each memory stores a metadata object containing the classified type, the active git branch, a confidence of 1.0, and the source set to remember_command. Confidence is 1.0 because the user explicitly requested the storage.

What happens if no text is provided to the remember command?▼

The Skill asks the user what should be remembered instead of failing. Once the user supplies the content, it proceeds with classification, storage, and confirmation as normal.