Make Claude Code Know You

Your AI is dumb because it knows nothing about you. Same model everyone else uses. Same generic answers. Now compare that to mine. Every task I write, every idea from a walk, every journal entry, every dollar I spend this week, it all sits in one place that Claude reads before it does anything. Tasks I wrote at 7 AM, it knows. Last week's wins and losses, it knows. The output is night and day better. It knows me for me, not a generic version of me.

That graph is my actual brain. Every dot is a file. Every line is a connection. Claude reads all of it. The setup below is how you build the same thing for yourself in about 10 minutes. Free. Your files stay on your computer. Nobody else touches them.

If you got that working you are already ahead of most people who try this. Now grab the starter folders below so your vault has a structure to actually grow into. Drop the folders into the vault you just created and you are good.

I just created an Obsidian vault at [PASTE YOUR VAULT PATH HERE]. Help me set up a simple daily log system inside it.

1. Create a folder called daily-log inside the vault.
2. Build a daily note template with short sections for what I did today, what I am thinking about, ideas worth saving, and anything on my mind.
3. Show me how to open a fresh daily note in under 10 seconds so I actually stick with it.

Keep it simple. No plugins, no fancy setup. I want to be logging within 2 minutes.

Run that prompt and Claude builds your daily log structure in front of you. That is your first win. Now the part most people miss. An Obsidian vault alone is just a folder of notes. The unlock is pointing Claude Code at it so every conversation starts with full context on you.

Update my global CLAUDE.md file at ~/.claude/CLAUDE.md to include my Obsidian vault as a personal context source. My vault lives at [PASTE YOUR VAULT PATH HERE].

Add a clear section that tells you:
1. Where my vault is on disk
2. What is inside it (daily logs, ideas, notes, thoughts, whatever folders I have)
3. That you can read from this vault anytime I reference my notes, my journal, my thoughts, my brain, or my personal context

Keep the new section short and direct. After you save the file, confirm which projects this will apply to and tell me exactly how you will know when to check the vault.

Now the good stuff. Every conversation with Claude Code starts with your full context. You do not need to re-explain who you are, what you are working on, or what you care about. Claude just checks your notes and knows. Run the prompt below in a fresh chat and watch what happens.

Read my Obsidian vault from the past 2 weeks. Pull out the themes I have been writing about, the ideas I have had, and anything I seem stuck on. Then suggest one thing I should act on today based on what is actually in my head, not what you think I should do.

Be specific. Reference the actual notes you saw. Do not make anything up.

That right there is the moment that hooks people. Watching AI pull from your actual notes and reflect something back you needed to hear. The first time it happens you stop using generic AI chats forever.

A few quick things to know. Your vault is just a folder of markdown files. Nothing fancy. If something ever looks broken, open the folder in Finder and you can see every note as a plain text file. You own everything.

Be thoughtful about what you write. Claude Code can read your entire vault, which means anything sensitive (passwords, financial info, private details you do not want in an AI's context) should live somewhere else. Keep the vault for notes and ideas, not secrets.

That is the whole foundation. Install Obsidian. Pick a folder. Start writing. Point Claude Code at it. Watch AI become something that actually knows you. Not even exaggerating.. once you experience this you do not go back.