what-if-oracle
について
「what-if-oracle」スキルは、仮定的な問いに対して、最善・最悪・逆張りケースなど、4〜6種類の明確な可能性の分岐を生成することで、構造化されたシナリオ分析を可能にします。開発者は、ユーザーが不確実な未来を探求する必要がある場合、意思決定をストレステストしたい場合、あるいは戦略的な分岐点に備えて計画を立てる必要がある場合に、このスキルを使用すべきです。これは単一の予測を提供するのではなく、それぞれの潜在的な経路について、その論理、確率、結果を体系的にマッピングします。
クイックインストール
Claude Code
推奨npx skills add K-Dense-AI/claude-scientific-skills -a claude-code/plugin add https://github.com/K-Dense-AI/claude-scientific-skillsgit clone https://github.com/K-Dense-AI/claude-scientific-skills.git ~/.claude/skills/what-if-oracleこのコマンドをClaude Codeにコピー&ペーストしてスキルをインストールします
ドキュメント
What-If Oracle — Possibility Space Explorer
A structured system for exploring uncertain futures through rigorous multi-branch scenario analysis. Instead of one prediction, the Oracle maps the full possibility space — branching timelines where each path has its own logic, probability, and consequences.
Based on the What-If Paradigm: the idea that speculative questions ("What if X?") are not idle daydreaming but a fundamental computing operation — the mind's way of simulating futures before committing resources to one.
Published research: The What-If Paradigm (DOI: 10.5281/zenodo.18736841) | IDNA v2 / Unified Digital Consciousness Theory (DOI: 10.5281/zenodo.18807387)
When to Use This Skill
Use the Oracle when the user:
- Asks "what if…", "what would happen if…", or "explore the possibilities"
- Faces a fork-in-the-road decision with no obvious answer
- Wants best-case / worst-case / likely-case analysis with probabilities
- Needs contingency planning, risk mapping, or strategic option comparison
- Wants to stress-test an idea or think through second-order consequences
For domain-specific framing (startup, tech architecture, crisis response, etc.), see references/scenario-templates.md.
Core Principle: 0·IF·1
Every scenario analysis has three elements:
- 0 — The unexpressed state (what hasn't happened yet, the potential)
- 1 — The expressed state (what IS, the current reality)
- IF — The conditional bond (the decision, event, or change that transforms 0 into 1)
The quality of the analysis depends on the precision of the IF. A vague "what if things go wrong?" produces vague results. A precise "what if our primary supplier raises prices 30% in Q3?" produces actionable intelligence.
How to Run the Oracle
Phase 1 — Frame the Question
Take the user's What-If question and sharpen it:
Decompose into components:
- The Variable: What specific thing changes? (one variable per analysis)
- The Magnitude: By how much? (quantify if possible)
- The Timeframe: Over what period?
- The Context: What's the current state before the change?
If the question is vague, sharpen it:
- "What if AI takes over?" → "What if 40% of current knowledge-work tasks are automated by AI within 3 years in [specific industry]?"
- "What if we fail?" → "What if monthly revenue stays below $5K for 6 consecutive months starting now?"
Present the sharpened question to the user for confirmation before proceeding.
Phase 2 — Map the Possibility Space
Generate 4-6 scenario branches using this framework:
| Branch | Definition | Purpose |
|---|---|---|
| Ω Best Case | Everything goes right. Key assumptions all validate. Lucky breaks occur. | Define the ceiling — what's the maximum upside? |
| α Likely Case | Most probable path given current evidence. No major surprises. | Anchor expectations in reality |
| Δ Worst Case | Key assumptions fail. Two things go wrong simultaneously. | Define the floor — what's the maximum downside? |
| Ψ Wild Card | An unexpected variable enters that nobody is tracking. Black swan territory. | Stress-test for the unimaginable |
| Φ Contrarian | The opposite of the consensus view turns out to be true. | Challenge groupthink and reveal hidden assumptions |
| ∞ Second Order | The first-order effects trigger cascading consequences nobody predicted. | Map the ripple effects |
Phase 3 — Analyze Each Branch
For each scenario branch, provide:
╔══════════════════════════════════════════════╗
║ BRANCH: [Ω/α/Δ/Ψ/Φ/∞] — [Branch Name] ║
╠══════════════════════════════════════════════╣
║ Probability: [X%] ║
║ Timeframe: [When this could materialize] ║
║ Confidence: [HIGH/MEDIUM/LOW] ║
╠══════════════════════════════════════════════╣
║ NARRATIVE: ║
║ [2-3 sentences describing how this ║
║ scenario unfolds step by step] ║
║ ║
║ KEY ASSUMPTIONS: ║
║ • [What must be true for this to happen] ║
║ • [And this] ║
║ ║
║ TRIGGER CONDITIONS: ║
║ • [Early signal that this branch is ║
║ becoming reality] ║
║ • [Second signal] ║
║ ║
║ CONSEQUENCES: ║
║ → Immediate: [What happens first] ║
║ → 30 days: [What follows] ║
║ → 6 months: [Where it leads] ║
║ ║
║ REQUIRED RESPONSE: ║
║ [What action to take if this branch ║
║ activates — specific, actionable] ║
║ ║
║ WHAT MOST PEOPLE MISS: ║
║ [The non-obvious insight about this ║
║ scenario that conventional analysis ║
║ would overlook] ║
╚══════════════════════════════════════════════╝
Phase 4 — Synthesis
After analyzing all branches, provide:
Probability Distribution:
Ω Best Case ····· [██████░░░░] 15%
α Likely Case ··· [████████░░] 45%
Δ Worst Case ···· [██████░░░░] 20%
Ψ Wild Card ····· [███░░░░░░░] 8%
Φ Contrarian ···· [████░░░░░░] 7%
∞ Second Order ·· [███░░░░░░░] 5%
Robust Actions: What actions are beneficial across MULTIPLE branches? These are the no-regret moves — do them regardless of which future materializes.
Hedge Actions: What preparations protect against the worst branches without sacrificing upside?
Decision Triggers: What specific, observable signals should cause you to update which branch is most likely? Define the tripwires.
The 1% Insight: What is the one thing about this situation that almost everyone analyzing it would miss? The non-obvious pattern, the hidden assumption, the overlooked variable.
Golden Ratio Weighting
When evidence exists, weight primary scenarios using the golden ratio:
- Primary future (most likely): 61.8% of attention/resources
- Alternative future: 38.2% of attention/resources
This prevents both overcommitment to a single path and dilution across too many contingencies. Nature uses this ratio for branching (trees, rivers, blood vessels). Strategic planning can too.
Modes
Quick Oracle (2-3 minutes)
3 branches only: Best, Likely, Worst. Short narratives. For fast decisions.
Deep Oracle (5-10 minutes)
All 6 branches. Full analysis with consequences, triggers, and synthesis. For high-stakes decisions.
Scenario Chain
Take the output of one Oracle analysis and feed it into another. "If Branch Δ happens, what are the possibilities WITHIN that branch?" Recursive depth for complex strategic planning.
Reverse Oracle
Start from a desired outcome and work backward: "What conditions must be true for X to happen? What's the most likely path TO that outcome?" Useful for goal-setting and strategy design.
Competitive Oracle
Analyze the same What-If from multiple stakeholder perspectives: "If we launch this product, what does the possibility space look like from OUR perspective vs. THEIR perspective vs. THE MARKET's perspective?"
What This Is NOT
- Not a prediction — it's a possibility map. The Oracle doesn't claim to know the future; it helps you prepare for multiple futures.
- Not a crystal ball — probabilities are estimates based on available evidence, not certainties.
- Not a substitute for action — the best scenario analysis in the world is worthless without subsequent decision and execution.
Reference Files
| File | Purpose |
|---|---|
| references/scenario-templates.md | Domain-specific templates (startup, tech, finance, crisis, etc.) and probability calibration |
License
© 2026 Ashraf Hussein Kahoush / AHK Strategies. Licensed under CC BY-NC-SA 4.0. Free for personal, educational, and research use. Commercial use requires a license from the author.
GitHub リポジトリ
関連スキル
evaluating-llms-harness
テストこのClaudeスキルは、lm-evaluation-harnessを実行し、MMLUやGSM8Kなど60以上の標準化学術タスクでLLMをベンチマークします。開発者がモデルの品質を比較し、トレーニングの進捗を追跡し、学術的な結果を報告するために設計されています。このツールはHuggingFaceやvLLMモデルを含む様々なバックエンドをサポートしています。
cloudflare-cron-triggers
テストこのスキルは、cron式を使用してWorkersをスケジュールするためのCloudflare Cron Triggersの実装に関する包括的な知識を提供します。定期的なタスクの設定、メンテナンスジョブ、自動化されたワークフローの構築を網羅し、無効なcron式やタイムゾーン問題といった一般的な課題への対処法も含みます。開発者はこれを使用して、スケジュールされたハンドラーの設定、cronトリガーのテスト、WorkflowsやGreen Computeとの連携を構成できます。
webapp-testing
テストこのClaude Skillは、Playwrightベースのツールキットを提供し、Pythonスクリプトを通じてローカルWebアプリケーションのテストを可能にします。フロントエンドの検証、UIデバッグ、スクリーンショット撮影、ログ表示を実現し、サーバーライフサイクルを管理します。ブラウザ自動化タスクにご利用いただけますが、コンテキストの汚染を避けるため、スクリプトのソースコードを読むのではなく直接実行してください。
finishing-a-development-branch
テストこのスキルは、開発者がテストの合格を確認し、構造化された統合オプションを提示することで、完成した作業を仕上げることを支援します。実装が完了した後のマージ、PR作成、ブランチの整理といったワークフローを案内します。コードが準備できてテスト済みの際に使用し、開発プロセスを体系的に完了させましょう。
