AI Memory Vault
A build script and templates that turn an Obsidian vault into an AI's working memory: markdown notes, no vector database
Medium risk
We rate an entry medium when the tool runs code, makes network calls or reads project files. Check what exactly it does before installing.
Why this level
- Builds a memory structure in your working environment and writes files
- May install Obsidian if needed
Install
Manual install
I'd like to set this up, please: https://github.com/jaredrhod/ai-memory-vault.git
Paste this into a Claude Code session; the ai-memory-vault.md build script builds the system with you.
This is third-party code. Review the repository files before installing.
What it does
AI Memory Vault turns an Obsidian vault into persistent AI memory in plain markdown, with no vector database. The centerpiece is the build script ai-memory-vault.md: you run it inside Claude Code, it checks whether Obsidian is installed and installs it if needed, interviews you and assembles a whole system: a boot config, a folder structure around your projects, daily notes, a profile that grows as it learns and jobs that teach the AI your recurring tasks. It ships templates CLAUDE.md, VAULT-INDEX.md, DAILY-NOTE.md and MEMORY.md with fill-in spots. The priming idea: before a task the AI reads the right set of your notes and then answers.
Who it is for. For people working with Claude Code who want persistent memory built on their own Obsidian notes.
Good fit when
- You want AI memory in plain markdown with no vector database
- You want the AI to read your notes before a task
- You want a system with a profile and daily notes around your projects
Not a fit when
- You do not use Obsidian and do not want to install it
- You need semantic search over a large corpus rather than markdown notes
Example request
Set up an Obsidian memory system from this repository and build the boot config and folder structureLimitations
The main path is built and tested on Claude Code; other terminal AIs that read and write files work with rough edges. Memory is markdown, with no semantic search or vector database. The system is built inside your working environment and touches its own files; updates only touch the repo's files, and your notes and config stay yours. The license is CC BY-SA 4.0.
How to disable. Delete the cloned repository and, if you want, the files it built (the boot config and the vault); your Obsidian notes stay untouched.
Security check
- Builds a memory structure in your working environment and writes files
- May install Obsidian if needed
README in short
The README describes AI Memory Vault as an open system and templates that turn an Obsidian vault into markdown AI working memory, with no vector database. The build script ai-memory-vault.md runs inside Claude, checks and if needed installs Obsidian, interviews you and builds the system: a boot config, a folder structure, daily notes, a profile and jobs for recurring tasks. The priming idea is that the AI reads the right set of your notes before its output. It ships templates CLAUDE.md, VAULT-INDEX.md, DAILY-NOTE.md and MEMORY.md, and updates only touch the repo's files. The license is CC BY-SA 4.0.
FAQ
Is a vector database needed?
No. Memory is plain markdown in an Obsidian vault, with no vector database or external service.
Will an update overwrite my notes?
No. Updates only touch the repo's files; your notes and CLAUDE.md live outside the repo and stay yours.
Related
A self-hosted knowledge base with block-level references and a built-in MCP server for connecting AI agents to your notes
A CLI for every Google Workspace API with JSON output and agent skills: Drive, Gmail, Calendar, Sheets and more
Local search over Markdown notes, docs and meeting transcripts: keywords, semantic search and reranking, with an MCP server
A task manager for AI-driven development: breaks a PRD into dependent tasks and guides the agent through them via MCP or CLI