OferBlum

Personal Digital Brain With AI — AI skill for Claude Code

AI community

An agentic Obsidian vault, split into six small repos that each show one idea: layout, privacy guard, frontmatter standard, LLM wiki, semantic search, Hermes layer.

How to install Personal Digital Brain With AI

This entry records only its repository, not the path inside it, so there is no exact command to give. Open OferBlum/personal-digital-brain-with-ai and copy the folder into ~/.claude/skills/, or the file into ~/.claude/agents/.

What Personal Digital Brain With AI does

An agentic Obsidian vault, split into six small repos that each show one idea: layout, privacy guard, frontmatter standard, LLM wiki, semantic search, Hermes layer. Start here.

Alternatives in AI

  • AGENTS.md - Skill Seekers — Concise reference for AI coding agents 11.1k ★
  • Pebrel — AI-native, GPU-accelerated terminal emulator for Windows with SSH, persistent sessions, split panes, and first 670 ★
  • Open Researcher — 🔥 Visual AI research assistant that displays real-time thinking, provides split-view analysis, and automatic 634 ★

README

Digital Brain — An Agentic Obsidian Vault

The biggest advantage of AI is also its weakness: it knows everything, so its answers don't always fit **you**. My fix is a knowledge base with all of **my** data — my Obsidian vault. AI agents get access to it and become an assistant that knows about me.

Technically, it's a **RAG** (retrieval-augmented generation) system. Before an agent answers, it finds the notes in my vault that fit the question and answers from them, not just from what the model already knows.

**I write the ideas, the agents read and enrich.**

What it does:

  • A wiki of summaries of everything I watch and read
  • Search by meaning, in Hebrew and English
  • A private folder that no agent can reach
  • Any agent — the rules, skills and memory live in the vault, not in one AI tool

The pieces

This repo used to hold the whole system, but it got too long to read. So I split it into six small repos. Each one shows one idea, and you can copy it on its own:

Repo What it does
agentic-vault-skeleton The layout. One AGENTS.md for every agent, a three-file map so agents don't scan the whole vault, skills that load only when needed, and a session cache
vault-privacy-guard Privacy is a folder, not a flag. One policy file, and a hook that checks every tool call in every agent
okf-frontmatter-standard The metadata on top of each note: three closed fields (type, subject, tags), plus a mark for what the AI wrote and what I checked
llm-wiki-youtube-transcript The Karpathy method: an agent turns each source (YouTube, articles, PDFs) into a linked wiki page
vault-semantic-search The retrieval part of the RAG. Find no