body{font-family:-apple-system,BlinkMacSystemFont,'Segoe UI',sans-serif;max-width:850px;margin:0 auto;padding:2rem;line-height:1.7;color:#e0e0e0;background:#0a0a0a}h1{font-size:2rem;border-bottom:2px solid #3b82f6;padding-bottom:.5rem;color:#fff}h2{font-size:1.4rem;color:#3b82f6;margin-top:2rem}.fw{background:#111827;border:1px solid #1e3a5f;border-radius:12px;padding:1.25rem;margin:1.25rem 0}.fw h3{margin:0 0 .5rem;color:#fff;font-size:1.1rem}.fw-p{color:#f59e0b;font-weight:600;font-size:.95rem}.usage{background:#0d1117;border-left:3px solid #3b82f6;padding:.5rem 1rem;margin:.5rem 0;border-radius:0 6px 6px 0;font-size:.9rem}table{width:100%;border-collapse:collapse;margin:1rem 0}th,td{padding:.5rem;text-align:left;border-bottom:1px solid #2a2a3e;font-size:.9rem}th{background:#1a1a2e;color:#3b82f6}a{color:#60a5fa}pre{background:#1a1a2e;padding:1rem;border-radius:8px;border-left:3px solid #3b82f6;font-size:.9rem;color:#a5b4fc}

Part of the Forge Neural Registry

12 Thinking Frameworks

Codified from 3 months of building Forge. Each is a repeatable decision tool. Combined, they form the governance layer of how Jason thinks, builds, and operates.

1. Domino Thinking

Find the 4% that produces 64%. Not all tasks are equal.

Prompt: "What's the ONE thing that makes 3+ others unnecessary?"

2. Abstraction Climbing

Never fix at the same level twice. Trace UP to the source.

Prompt: "What CREATES the input that fails? Fix the creator, not the cleanup."

3. Core Truth Anchoring

5-line vision anchor: first principle, problem, desired cascade, anti-goals, test.

Prompt: "WHY does this exist? What pain? What 2nd-order effects? What must it NOT become? How do we know it worked?"

4. Decision Gate (6 Binary Checks)

Not scored 1-5. Six yes/no questions.

Prompt: "Domino? Cascade 3 steps? Root cause clear? Breaks if skip 30d? Enough info? Reversible?"

5. Antifragile Loop (3-Strike)

Strike 1: patch. Strike 2: trace up. Strike 3: redesign. Strike 5: halt + publish.

Prompt: "Which strike am I on? Am I fixing at the same level as last time?"

6. Surface Everything

If the user can't see it, it doesn't exist.

Prompt: "Can Jason find this on his phone in 30 seconds?"

7. Right-Size It

V1 dirty. V2 solid. V3 never.

Prompt: "Am I accidentally proposing V3? What's the cheapest proof this matters?"

8. Cascade Mapping

Nothing is binary. Trace 3 steps forward. Invert: what breaks if you don't?

Prompt: "[Action] → enables [?] → compounds into [?]. Skip 30 days: what breaks?"

9. No Dead Ends

Every failure owned within 30 minutes. Retry, recover, or escalate.

Prompt: "Is anything stuck that nobody owns? That's the first thing to fix."

10. Fix at Source

A factory making defective parts needs to fix the machine, not hire inspectors.

Prompt: "I've fixed this twice at level X. What's above X that CREATES the problem?"

11. Smoke Test Before Pipeline

Check the simple proof points FIRST. Don't build an elaborate system to find out step 1 was wrong.

Prompt: "What's the cheapest check that proves this is even the right problem?"

12. Meta-Climb

Don't solve at the level it presents. Go up until one action eliminates a CLASS.

Prompt: "Is there a level above where I solve ALL instances with one move?"

How They Combine

NEW WORK → Core Truth → Domino → Decision Gate → Right-Size → Cascade → Build → Surface → Smoke Test

FAILURE → Antifragile Loop → Abstraction Climbing → Fix at Source → Meta-Climb → No Dead Ends

As Expert Clone Prompts

SituationLoad These
"What should I work on?"Domino + Decision Gate
"This keeps breaking"Antifragile + Abstraction Climbing + Fix at Source
"Should we build this?"Core Truth + Cascade + Right-Size
"Something failed"No Dead Ends + Smoke Test + Antifragile
"Is this done?"Surface Everything + Core Truth

← Back to Neural Registry | ← How Forge Heals Itself

]]>