jaredrhod/ai-memory-vault

What it solves

It solves the problem of AI "forgetfulness" and context limits by providing a persistent, external memory system. Instead of relying on the model's limited session memory, it allows an AI (specifically designed for Claude Code or similar terminal AIs) to store and retrieve information from a permanent knowledge base.

How it works

The system turns an Obsidian vault into the AI's working memory. It uses a build script to set up a folder structure, boot configurations, and templates. The AI is configured to read specific notes—a process called "AI Priming"—before performing tasks, ensuring it has the necessary context (such as company knowledge or personal preferences) to produce accurate results. It also redirects the AI's native memory back into the vault to prevent data fragmentation.

Who it’s for

It is designed for users of Claude Code or other terminal-based AI agents who want a long-term, scalable memory system that lives outside the AI model and can be managed manually via Obsidian.

Highlights

• Obsidian Integration: Uses a standard Obsidian vault as the primary storage for AI memory.
• AI Priming: Allows the AI to read specific sets of notes tailored to a particular "Job" before generating output.
• Boot Configuration: Includes a pre-configured agent identity (personality and rules) that can be used as-is or customized.
• No Size Ceiling: Because memory is stored in local files, it is not limited by the AI model's context window.