skills.rest
semicolon-devteam
semicolon-devteamOFFICIAL
Explore all skills by this author →
WebsiteGitHub

memory

Persist SEMO session context across sessions using versioned JSON schemas.

Updated Dec 11, 2025
Search Tags:#automation#session management#AI memory#architectural decisions#SEMO#context persistence#team preferences
One-click install
npx skills add https://github.com/semicolon-devteam/semo --skill memory-semicolon-devteam
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: memory
Source: https://github.com/semicolon-devteam/semo/tree/main/semo-skills/memory
Command: npx skills add https://github.com/semicolon-devteam/semo --skill memory-semicolon-devteam

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill requires Read, Write, Bash, jq, and includes references (resource) components.

What problem does it solve?

This Skill addresses the "amnesia" problem in AI agents by providing a long-term memory system that persists architectural decisions, team preferences, and project context across sessions, ensuring consistency and reducing repetitive instructions.

Core Features & Use Cases

  • Decision Storage: Saves architectural decisions (e.g., API patterns, error handling strategies) for future reference.
  • Preference Recording: Stores team or user preferences (e.g., code style, naming conventions) to apply consistently.
  • Context Synchronization: Automatically loads saved decisions, preferences, and project context at the start of each session.
  • MCP Integration: Supports integration with Memory MCP servers for more robust and shared memory persistence.
  • Use Case: After deciding on a new API response pattern, you can instruct "skill:memory save decision 'api-pattern' '모든 API 응답은 JSON Envelope 패턴 사용'" to ensure all future API implementations adhere to this decision.

Quick Start

Save a decision: "skill:memory save decision 'api-pattern' 'JSON Envelope 패턴 사용'" Load all decisions: "skill:memory load decision" Sync memory at session start: "skill:memory sync"

Frequently Asked Questions about memory

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

FAQPage Schema
How do I persist architectural decisions across AI sessions?▼

Persist architectural decisions by saving them to a .claude/memory directory structure using the save command. The Skill stores decisions in decisions.json, automatically loads them at session start, and applies them consistently across future sessions without repetitive instructions.

Can I store and apply team preferences automatically?▼

Yes. Store team preferences like code style and naming conventions in preferences.json within the memory directory. The Skill loads these preferences automatically on session start and applies them consistently to all subsequent work.

What directory structure do I need to set up for memory?▼

Set up a .claude/memory directory with subdirectories: decisions.json, preferences.json, cache/, and context/{project}.json. This structure enables the Skill to organize and retrieve architectural decisions, team preferences, cached data, and project-specific context across sessions.

How does the Skill integrate with Memory MCP servers?▼

The Skill integrates with Memory MCP when available for more robust and shared memory persistence. It uses versioned JSON schemas to ensure consistent migration and compatibility with MCP-based memory backends for team-wide context synchronization.

What commands can I use to manage saved decisions and preferences?▼

Use three core commands: save to store decisions and preferences, load to retrieve them, and sync to reload all memory at session start. Each command operates on decisions.json and preferences.json files within the .claude/memory directory.

Does this work if I'm switching between different projects?▼

Yes. The Skill organizes project-specific context in context/{project}.json files within the memory directory. You can sync memory for different projects, and the appropriate context loads automatically based on the active project.

Related Skills

Explore More Software Engineering Skills→
teaglebuiltteaglebuilt

init-memory

Create context.json, sessions.json, and decisions.json for project memory.

Community
Basic
ajayatwal1105-emersonajayatwal1105-emerson

sync-memory

Sync new and modified marketing learnings into persistent memory for an active brand.

Community
Advanced
arunrlvermaarunrlverma

agent_memory

Store and retrieve AI agent facts, decisions, and preferences in a cloud JSON file.

Community
Intermediate
zach-sourcezach-source

graphiti

Store and retrieve persistent AI memory in a FalkorDB temporal knowledge graph.

Community
Advanced