kortix-memory

Manage persistent user preferences and per-project context via USER.md, MEMORY.md, and CONTEXT.md files.

20.1k|3.4k|Updated Oct 5, 2024
One-click install
npx skills add https://github.com/kortix-ai/suna --skill kortix-memory
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: kortix-memory
Source: https://github.com/kortix-ai/suna/tree/main/core/kortix-master/opencode/skills/KORTIX-system/memory
Command: npx skills add https://github.com/kortix-ai/suna --skill kortix-memory

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Kortix memory provides a lightweight, persistent memory layer that injects global user preferences and per-project context into every chat turn, reducing repetitive setup and improving consistency across sessions.

Core Features & Use Cases

  • Global memory injection: USER.md for user preferences and MEMORY.md for global constants and conventions.
  • Per-project CONTEXT.md injection: CONTEXT.md is injected automatically when the session is linked to a specific project.
  • CRUD-style memory management: read, edit, and write memory files to update knowledge without rewriting entire files.
  • Token-efficient design: top-level memory is concise with deeper notes stored in subfiles.
  • Use cases: maintain long-running workflows across multiple projects and ensure consistent agent behavior.

Quick Start

Turn on memory by linking a session to a project to activate automated memory injection of USER.md, MEMORY.md, and CONTEXT.md for seamless, context-aware interactions.

Frequently Asked Questions about kortix-memory

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

FAQPage Schema
How do I persist user preferences and project context for AI agents across multiple sessions?

You can persist AI agent context using a file-based memory schema with USER.md for global preferences, MEMORY.md for constants, and CONTEXT.md for project-specific details injected into every chat turn.

What is the best way to manage AI agent memory without exceeding token limits?

Token-efficient memory management is achieved by keeping top-level memory files concise while storing deeper notes in subfiles, ensuring persistent context injection does not consume excessive tokens during chat sessions.

How does per-project context injection work for multi-project AI workflows?

Per-project context injection works by automatically reading a CONTEXT.md file when a session is linked to a project, ensuring the AI agent receives dynamic, project-specific memory updates for consistent behavior.

Can I update persistent AI memory files without rewriting the entire context?

Yes, you can update persistent AI memory files using CRUD-style memory management, allowing you to read, edit, and write specific memory entries to update knowledge without rewriting entire files.

Do I need a database for persistent file-based memory management in AI agents?

No database is needed for file-based memory management; the system uses flat files like USER.md, MEMORY.md, and CONTEXT.md to store global preferences and project context across multiple sessions.

Why does my AI agent lose user preferences and consistency across different chat sessions?

AI agents lose consistency without persistent memory injection; linking a session to a project activates automated injection of memory files to maintain dynamic updates and consistent behavior across workflows.