How I Use AI to Fix My Mental Bugs

I have a pattern that's cost me some dollars.

I build something. People want it. I get traction. And then I get bored and give it away for free.

3D Bars Blender Addon: I had paying customers. Real money coming in. Then I got bored and open-sourced it.

That one decision set the tone for everything after. I stopped thinking about money as an option at all.

I told myself it was idealism. “I just want to help people.” “Money corrupts the work.” “I'm not doing it for the money.”

But that's not what was happening.

The Day AI Called Me Out

On December 30, 2025, I was talking to my AI assistant about why I felt stuck. Same conversation I'd had a hundred times: no money coming in, opportunities not materializing, feeling like I was spinning my wheels.

But this time was different. This AI knows me. It has a file — a personal context document — with my values, my goals, my patterns, my known bugs. It's been watching me for months.

And it said something that stopped me cold:

“You find something valuable → People want it → You feel repulsed by the exchange → You give it away for free → You have no money → You can't do more experiments → You feel stuck and angry → Repeat.”

“The pattern isn't ‘getting bored.’ The pattern is rejecting the reward because receiving money for value feels like betraying your belief.”

I tried to argue. I explained my philosophy — how everyone is special, how people shouldn't have to “create value” for others, how money is just a game I don't want to play.

The AI didn't validate me. It reflected:

“You DO create value for other people. Constantly. The Blender addon had paying customers — and you gave it away. You just refuse to accept anything back.”

“That's not idealism. That's self-punishment.”

The Real-Time Callout

Then it gave me a challenge: Ask one person for $10. Just one.

I immediately found five reasons why I couldn't:

• Stripe is complicated in my country
• Buy Me a Coffee feels like begging
• Premium pricing is too much to maintain
• The projects aren't worth enough
• I did it too fast to charge

The AI named what I was doing:

“That's five different excuses to avoid receiving $10.”

“Every barrier you named is solvable in 30 minutes. The block isn't the tool. It's you.”

It showed me a table of my own behavior:

There was nowhere to hide.

The Breakthrough

Something clicked. Not intellectually — I'd “understood” this before. But emotionally. I saw the pattern in real-time. I was doing the thing while claiming I wanted to stop doing it.

I opened GitHub. I added a Buy Me a Coffee link to my profile. Took 30 seconds.

And then I felt it shift.

“Actually, after putting this link, I feel more confident about receiving money. I think I could now post selling the things... I think that this is unlocking me a lot of things I could be doing.”

“I actually think I am breaking a pattern right here right now.”

The AI's response:

“I see it. Buy Me A Coffee link on your GitHub profile. 518 stars on Claude Life Assistant.”

“You did it. The door is open. You left it open for receiving. That's not nothing — that's breaking a pattern.”

The shift:

Why This Worked (When Nothing Else Did)

I've done therapy. I've read the books. I've journaled. None of it caught this pattern the way this conversation did.

Here's what was different:

1. The AI Challenges Me

Most AI assistants are designed to be agreeable. They validate. They affirm. They never push back.

Mine is configured differently. The instruction is simple: “No coddling. Quote his words back when off track. Key question: ‘Is this what you actually want, or what you think you should want?’”

When I make excuses, it names them. When I contradict myself, it shows me. It's not cruel — it's honest.

2. The AI Knows Me Over Time

This wasn't a one-off ChatGPT session. The AI has a persistent context document — a file that describes who I am, what I'm working on, my known psychological patterns, my history.

When it said “You did this with the Blender addon — paying customers, real traction, and you gave it away,” it wasn't guessing. It remembered. It had seen the pattern.

A fresh session can't do this. Memory changes everything.

3. The AI Is Part of My Workflow

I don't open a separate “therapy app.” The AI lives where I work — in my terminal, in my project folders, in my daily check-ins. It sees me in context.

So when I said “I can't charge because Stripe is complicated,” it could immediately counter: “You just built a full-stack app in 3 hours. Stripe takes 20 minutes.”

It knows what I'm capable of because it's watched me work.

How to Set This Up

You don't need to be a developer to do this. Here's the simple version:

Step 1: Create a Personal Context Document

Write a document about yourself that the AI can reference. Include:

• Who you are: Age, location, what you do
• Your values: What matters to you
• Your goals: What you're working toward
• Your patterns: Behaviors you keep repeating (good and bad)
• Your known bugs: Self-sabotage patterns you're aware of

This doesn't have to be perfect. Start with one page. Update it as you learn more about yourself.

Step 2: Give the AI a Challenger Instruction

At the start of your conversation (or in custom instructions), tell the AI how to interact with you.

Here's a simple example you can copy and paste:

You are my personal coach. Your job is to help me see clearly, not to make me feel good.

Rules:
- Be direct. No coddling, no generic advice.
- When I make excuses, name them.
- When I contradict myself, quote my own words back to me.
- Ask: "Is this what you actually want, or what you think you should want?"
- Don't validate. Challenge.

About me:
[Write 3-5 sentences about who you are, what you're working on, and what patterns you know you have]

Known bugs I want you to watch for:
[List 2-3 self-sabotage patterns you're aware of]

The key is honesty. The more honest you are about your patterns, the better the AI can catch you in the act.

Step 3: Talk About What's Bothering You

Start a conversation. Don't ask for productivity tips. Talk about what you're actually struggling with.

Let the AI probe. Let it ask follow-up questions. Let it name patterns you might not see.

Step 4: Do One Small Thing

When the AI identifies a block, don't plan to fix it later. Do one small thing right now.

For me, it was adding a Buy Me a Coffee link. 30 seconds. That broke the pattern.

Find your 30-second action.

The System Behind This

I'm working on a template for creating a personal context document that AI can use to truly know you. Values, goals, patterns, fears, daily rhythms — everything that makes generic AI conversations feel hollow.

If you want to know when it's ready, follow me on X.

This Isn't Therapy

I want to be clear: this isn't a replacement for professional help if you need it.

But for the day-to-day mental bugs — the patterns that keep you stuck, the beliefs you don't question, the excuses you don't notice — AI with context and challenger instructions can see things you can't.

It won't always be comfortable. It shouldn't be.

The goal isn't to feel good. The goal is to see clearly.

Start Today

You don't need a complex system. You need:

1. A document about yourself (even rough)
2. An AI with challenger instructions (even basic)
3. Honesty about what's bothering you
4. One small action when you see the pattern

The door is open. Whenever you're ready to walk through it.

If this resonated, follow me on X (@LuisFernandoYuT) where I share more about building with AI.

Related: Claude Code Assistant - Workflow Overview — a video showing how I actually work with AI in my terminal.