memory

Store and retrieve persistent user context across sessions using markdown memory files.

Updated Aug 23, 2026

Search Tags:#context #memory #profiling #persistence #session-management #ai-assistant #long-term-memory

Ready to run

npx skills add https://github.com/cris-m/flopsy --skill memory-cris-m

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: memory
Source: https://github.com/cris-m/flopsy/tree/main/src/team/templates/skills/memory
Command: npx skills add https://github.com/cris-m/flopsy --skill memory-cris-m

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Working memory and contextual persistence for AI agents is often lost between sessions. This Skill provides a structured way to persist user context, profile preferences, and topic history so the agent can maintain continuity.

Core Features & Use Cases

Memory architecture with USER.md for profile, MEMORY.md for long-term decisions, and daily memory in memory/YYYY-MM-DD.md.
Topic tracking and heartbeat-ready memory to avoid repeating topics and to surface relevant context.
End-of-conversation consolidation that updates durable memory and profiles while preventing direct edits to MEMORY.md.
Use Cases: maintaining user preferences across sessions, long-running chats, and adaptive interactions based on history.

Quick Start

Remember the user’s preferences and recent topics in memory for continuity.

Frequently Asked Questions about memory

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

FAQPage Schema

How do I persist context across sessions for an AI assistant?▼

To persist context across sessions, you use a memory architecture that stores user profiles and long-term decisions in markdown files like USER.md and MEMORY.md, while tracking daily interactions in dated logs. This ensures continuity for long-running conversations.

What is the best way to maintain long-term memory in long-running chats?▼

The best way to maintain long-term memory is implementing a structured file system where MEMORY.md holds durable decisions and USER.md stores preferences. End-of-conversation consolidation updates these profiles to provide adaptive interactions based on history.

How does memory consolidation work for tracking user preferences across sessions?▼

Memory consolidation for tracking preferences works by capturing session data in a working file, then updating durable files like USER.md and MEMORY.md at the end of a conversation. Direct edits to MEMORY.md are prevented to protect data integrity.

Can I use markdown files for user profiling and topic tracking in AI agents?▼

Yes, you can use markdown files for user profiling and topic tracking by structuring data into USER.md for profiles and dated daily logs for topics. This heartbeat-ready memory approach surfaces relevant context and avoids repeating topics.

Why does an AI assistant lose working memory between conversations without session management?▼

An AI assistant loses working memory between conversations without session management because contextual persistence is not natively retained. Without a file-based memory architecture, user preferences and topic history are lost when the session ends.

When should I avoid directly editing durable memory files in session management?▼

You should avoid directly editing durable memory files like MEMORY.md during active sessions to prevent unconsolidated or fragmented data. Instead, use a working memory file for live updates and rely on end-of-conversation consolidation rules.