remember

Save user learnings to persistent .context/MEMORY.md files with optional scope.

Updated Jan 4, 2026
One-click install
npx skills add https://github.com/c-daly/agent-swarm --skill remember
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: remember
Source: https://github.com/c-daly/agent-swarm/tree/main/skills/remember
Command: npx skills add https://github.com/c-daly/agent-swarm --skill remember

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Saving user learnings and preferences across tools and sessions to improve Claude interactions and recall. This reduces repetitive prompts and manual reconfiguration.

Core Features & Use Cases

  • Automatic scope inference to store memory at user, component, or repository level
  • Persist memory to dedicated files like .context/MEMORY.md and confirm where memory was saved
  • Support remembering patterns, preferences, and learned behaviors across sessions
  • Use case: Remember user preferences or essential knowledge to tailor responses

Quick Start

Use the remember skill to save a preference or learning item, for example: /remember "I prefer concise outputs" or /remember --scope=project "Use agent-swarm remember feature for this project"

Frequently Asked Questions about remember

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

FAQPage Schema
How do I persist learnings and preferences across Claude sessions?

Persist learnings by saving them to memory files using automatic scope inference. The remember Skill saves your preferences and patterns to .context/MEMORY.md at user, component, or repository level, so Claude recalls them across sessions without repetition.

What does scope inference mean when saving memory?

Scope inference automatically detects whether your learning applies locally, to a component, or across a repository based on content and context. This ensures memory is stored at the right level—user-wide for global preferences, project-level for repository patterns—without manual specification.

How do I control where my memory gets saved?

Use the --scope parameter to explicitly set storage level: omit it for automatic inference, or specify --scope=project to force repository-level storage. Memory writes to .context/MEMORY.md with confirmation of the scope used.

Can I remember specific preferences to customize Claude's responses?

Yes. Save preferences like output style or domain knowledge using the remember Skill. Claude retrieves stored learnings across sessions, reducing the need to restate preferences and enabling tailored interactions based on your documented patterns.

What are the prerequisites for using the remember Skill?

The remember Skill requires Python and file IO capabilities to write memory files. No other dependencies are needed; memory persists to plain-text .context/MEMORY.md files accessible across sessions.

Does remember work across multiple projects or just one?

Remember supports multiple scope levels: user-level memory persists across all projects, while project-scoped memory stays within a repository. Automatic scope inference determines storage location based on your input context.