mindfuck 1.0.0

Brutal 5-perspective analysis for any idea, decision, or plan. No vague answers.

mindfuck

Not 30 models fighting — one framework that forces the question your brain avoids.

Drop any idea, decision, or plan. Get 5 brutally honest perspectives. No vague answers. No "it depends." No sugar coating. Just the truth from every angle.

Use It As a Claude Skill

Drop mindfuck.md into your Claude custom instructions, Cursor rules, or any AI tool that supports system prompts. Zero install. Works everywhere.

Trigger it by saying "No gay fuck" or asking for multi-perspective analysis on any question.

You: No gay fuck — should I quit my job to build this startup?

Claude:

## The Contrarian
Your runway is 6 months. That's not enough time to find PMF...

## First Principles Thinker
You're asking "should I quit?" but the real question is
"why can't I validate this while employed?"...

## The Expansionist
You're thinking about this one startup, but what if this
is actually your entry point into a bigger space?...

## The Outsider
I'm confused — why is quitting the only option?...

## The Executor
Here's what you do: Don't quit yet. This week, build a
landing page. Weekend: Run $200 in ads...

---

Synthesis: All 5 agree — you haven't validated demand yet.

One concrete next step:
This weekend: Build a landing page. Monday: Run it.

The 5 Perspectives

Perspective	What It Does
✗ The Contrarian	Finds the fatal flaw. Assumes you'll fail. Starts with "This will fail because..."
◎ First Principles	Strips assumptions. Reframes the whole question. "You're asking X, but the real question is..."
↗ The Expansionist	Finds the 10x version you can't see. "You're thinking too small..."
? The Outsider	Zero context. Pure fresh eyes. "I don't understand why..."
→ The Executor	What happens Monday morning. Concrete steps. "Here's what you do:"

Also Available as a CLI

pip install mindfuck

export ANTHROPIC_API_KEY=your_key_here

# analyze any idea
mindfuck "should I quit my job to go all-in on my startup?"

# pipe it
echo "my side project idea" | mindfuck

# raw JSON output
mindfuck "my idea" --raw

Why

Most AI tools tell you what you want to hear. This one doesn't.

Five named roles force the model into distinct cognitive stances — not 30 models running in parallel, but one framework that catches blind spots by design. The Contrarian finds what will break. First Principles reframes what you're actually solving. The Expansionist shows you the bigger play. The Outsider asks the dumb questions that aren't dumb. The Executor tells you what to do Monday morning.

It's the framework I use personally to cut through my own bullshit before making any real decision.

Requirements

• Skill: Any Claude-compatible AI tool (Claude.ai, Cursor, Antigravity, API)
• CLI: Python 3.9+ and ANTHROPIC_API_KEY

License

MIT