home / skills / danielmiessler / personal_ai_infrastructure / thinking

This skill stress-tests ideas and strategies across 11 world horizons, updating models and surfacing insights to guide high-stakes decisions.

npx playbooks add skill danielmiessler/personal_ai_infrastructure --skill thinking

Review the files below or copy the command above to add this skill to your agents.

Files (53)
SKILL.md
1.8 KB
---
name: WorldThreatModelHarness
description: Stress-test ideas, strategies, and investments across 11 time horizons (6mo-50yr). Update and view world models. USE WHEN threat model, world model, test idea, test strategy, future analysis, test investment, test against future, stress test idea, time horizon analysis, update models, view models, refresh models, model status.
---

# World Threat Model Harness

A system of 11 persistent world models spanning 6 months to 50 years. Each model is a deep (~10 page)
analysis of geopolitics, technology, economics, society, environment, security, and wildcards for that
time horizon. Ideas, strategies, and investments are tested against ALL horizons simultaneously using
adversarial analysis (RedTeam, FirstPrinciples, Council).

## Workflow Routing

| Trigger | Workflow | Description |
|---------|----------|-------------|
| "test idea", "test strategy", "test investment", "how will this hold up", "stress test", "test against future" | `Workflows/TestIdea.md` | Test any input against all 11 world models |
| "update world model", "update models", "refresh models", "new analysis" | `Workflows/UpdateModels.md` | Refresh world model content with new research/analysis |
| "view world model", "show models", "current models", "model status" | `Workflows/ViewModels.md` | Read and summarize current world model state |

## Tier System

All workflows support three execution tiers:

| Tier | Target Time | Strategy | When to Use |
|------|-------------|----------|-------------|
| **Fast** | ~2 min | Single agent synthesizes across all models | Quick gut-check, casual exploration |
| **Standard** | ~10 min | 11 parallel agents + RedTeam + FirstPrinciples | Most use cases, good depth/speed balance |
| **Deep** | Up to 1 hr | 11 parallel agents + per-horizon Research + RedTeam + Council + FirstPrinciples | High-stakes decisions, major investments |

**Default tier:** Standard. User specifies with "fast", "deep", or tier defaults to Standard.

## World Model Storage

Models are stored at: `~/.claude/MEMORY/RESEARCH/WorldModels/`

| File | Horizon |
|------|---------|
| `INDEX.md` | Summary of all models with last-updated dates |
| `6-month.md` | 6-month outlook |
| `1-year.md` | 1-year outlook |
| `2-year.md` | 2-year outlook |
| `3-year.md` | 3-year outlook |
| `5-year.md` | 5-year outlook |
| `7-year.md` | 7-year outlook |
| `10-year.md` | 10-year outlook |
| `15-year.md` | 15-year outlook |
| `20-year.md` | 20-year outlook |
| `30-year.md` | 30-year outlook |
| `50-year.md` | 50-year outlook |

## Context Files

| File | Purpose |
|------|---------|
| `ModelTemplate.md` | Template structure for world model documents |
| `OutputFormat.md` | Template for TestIdea results output |

## Skill Integrations

This skill orchestrates multiple PAI capabilities:

- **RedTeam** — Adversarial stress testing of ideas against each horizon
- **FirstPrinciples** — Decompose idea assumptions into hard/soft/assumption constraints
- **Council** — Multi-perspective debate on idea viability across horizons
- **Research** — Deep research for model creation and updates

## Voice Notification

Before any workflow execution:
```bash
curl -s -X POST http://localhost:8888/notify \
  -H "Content-Type: application/json" \
  -d '{"message": "Running WORKFLOW_NAME in the World Threat Model Harness", "voice_id": "fTtv3eikoepIosk8dTZ5"}'
```

## Customization Check

Before execution, check for user customizations at:
`~/.claude/PAI/USER/SKILLCUSTOMIZATIONS/WorldThreatModelHarness/`

Overview

This skill stress-tests ideas, strategies, and investments across 11 persistent world models spanning 6 months to 50 years. It runs adversarial analyses (RedTeam, FirstPrinciples, Council) and returns structured, horizon-by-horizon results so you can see how a plan performs across near and long-term futures. Use the default Standard tier for most runs, or choose Fast or Deep for quick checks or rigorous, high-stakes review.

How this skill works

The harness maintains 11 deep world models (6-month through 50-year) that capture geopolitics, technology, economics, society, environment, security, and wildcards for each horizon. When you submit an idea, strategy, or investment, the system runs parallel analyses across all horizons using configurable tiers: Fast (single-agent), Standard (parallel agents + RedTeam + FirstPrinciples), and Deep (full research + Council). Results are saved and can be refreshed when models are updated.

When to use it

  • Validate a business strategy or product roadmap against multiple future horizons
  • Stress-test an investment thesis for resilience across 6 months to 50 years
  • Run adversarial analysis on policy proposals, programs, or contingency plans
  • Refresh long-term planning when new research or geopolitical events occur
  • Get a quick gut-check (Fast) or deep, rigorous review (Deep) for high-stakes choices

Best practices

  • Choose the execution tier based on stakes: Fast for quick checks, Standard for routine analysis, Deep for major decisions
  • Provide concise assumptions and desired outcomes to improve FirstPrinciples decomposition
  • Review per-horizon recommendations and aggregated risks before deciding
  • Refresh world models periodically after major data updates or events
  • Use the stored model index to check last-updated dates before running tests

Example use cases

  • Test a five-year product pivot against technology and regulatory futures across all horizons
  • Evaluate a venture investment for durability under geopolitical and economic stressors
  • Assess a governmental policy for long-term social and environmental impacts
  • Compare competing strategies with adversarial red-teaming to surface failure modes
  • Schedule a model refresh and rerun tests after a disruptive global event

FAQ

What are the execution tiers and when should I use them?

Fast is a quick single-agent synthesis for informal checks. Standard is the default balance of depth and speed with parallel analysis and adversarial review. Deep adds per-horizon research, Council debate, and is for high-stakes decisions.

How often should I refresh the world models?

Refresh models after significant new research or major global events, or on a scheduled cadence that matches your decision rhythm (e.g., quarterly or after each major planning cycle).