Skip to content

Latest commit

 

History

History
294 lines (208 loc) · 10.3 KB

File metadata and controls

294 lines (208 loc) · 10.3 KB

🧠 CEOskill

Make AI think like a world-class CEO chief of staff: structure decisions, detect bias, predict competition, manage stakeholders.

License: MIT Skill Language

简体中文 | English


One Line

AI gives advice easily. But decomposing trade-offs, detecting cognitive bias, predicting competitive responses, and navigating stakeholder dynamics — that's the decision quality CEOs actually need.

5-Second Fit Check

If you face these problems regularly, this is built for you:

Scenario Typical Questions
🎯 Strategic Decisions Enter a new market? Build or acquire? Keep investing or cut losses?
💰 Resource Allocation 5 projects, budget for 2 — which ones? Growth vs product vs org?
🔥 Crisis Response CTO leaving for a competitor — what do you do in 2 hours?
🤝 Stakeholder Management How to align board, investors, and leadership? Manage reorg politics?
⚔️ Competitive War Gaming If we cut prices / launch a feature, how will competitors react?

The Problem: Why Generic AI Advisors Fall Short

Most "CEO advisor" AI prompts share the same weaknesses:

User: "Should we enter the Japan market?"

Generic AI:
"Entering the Japan market is an important strategic decision. You should
consider market size, competitive landscape, localization costs, and
regulatory requirements. I recommend conducting market research, assessing
risks, and developing a detailed plan. Would you like me to help you
identify some key questions to consider?"

❌ Sounds polished but stays vague
❌ Lists frameworks but doesn't choose the right one
❌ Gives advice without surfacing trade-offs
❌ Doesn't check cognitive bias (are you chasing FOMO because a competitor entered?)
❌ Doesn't predict competitive response (how will local players fight back?)
❌ Doesn't map stakeholders (who on your board supports vs opposes this?)

The Solution: How CEOskill Works

User: "Should we enter the Japan market?"

CEOskill:

【Decision Classification】Type 1 (One-Way Door) — High entry cost, high exit cost

【Bias Check】
- Are you experiencing FOMO because a competitor just entered?
- Are you overestimating first-mover advantage? (Japan rewards localization depth)
- Are you underestimating regulatory/cultural barriers?

【4 Options + Trade-offs】
A. Wholly-owned subsidiary ($5M, 18-month payback, risk: localization failure)
B. JV / acquire local player ($8M, 6-month launch, risk: integration conflict)
C. Lightweight pilot ($500K, validate PMF first, risk: miss the window)
D. Don't enter (save $5M+, focus current markets, risk: competitor builds moat)

【Competitive War Gaming】
- How will the top 3 local players respond? (Price war / channel lockout / regulatory lobbying)
- What does your competitor's performance look like 6 months after entry? (User growth vs burn rate)

【Stakeholder Map】
- Board: 2 support (bullish on international), 1 opposes (fears resource dilution)
- Investors: care about 18-month ARR growth, geography-agnostic
- Leadership: VP Sales supports, CFO neutral-leaning-against

【Recommendation】
Option C (pilot) + 6-month decision gate.
Rationale: Japan PMF uncertainty is high; going all-in is premature.
Next Steps:
1. Find 1-2 local design partners (2 weeks)
2. Run MVP to validate willingness to pay (3 months)
3. Data-driven decision on A/B/D (month 6)

✅ Structured decision workflow
✅ At least 3 comparable options + do nothing
✅ Explicit trade-offs
✅ Cognitive bias check
✅ Competitive + stakeholder response prediction
✅ Next steps + decision gate

CEOskill vs Generic AI Advisors

Capability Generic AI Consultant Prompt CEOskill
Decision classification (one-way / two-way / crisis) ⚠️ Sometimes ✅ Enforced
At least 3 options + do nothing ⚠️ Inconsistent ✅ Enforced
Explicit trade-offs (not just conclusions) ⚠️
Cognitive debiasing (FOMO / sunk cost / anchoring)
Competitive response prediction (war gaming) ⚠️
Stakeholder mapping (board / investors / leadership) ⚠️
Quantitative analysis (Monte Carlo / ICE / NPV)
Crisis / Prioritization / War Game modes

Core Capabilities

1. Structured Decision Workflow

Not "giving advice" — decomposing decisions:

  • Decision classification: One-Way Door / Two-Way Door / Crisis / Strategic Bet
  • Framework selection: Automatically picks the best framework for the scenario
  • Option generation: Forces at least 3 comparable options + do nothing
  • Trade-off analysis: Explicit cost, risk, and resource requirements for each option

2. Cognitive Debiasing

The biggest enemy of CEO decisions is not lack of information — it's cognitive bias:

  • Anchoring: Are you anchored to a number set by a competitor / investor / advisor?
  • Sunk cost: Are you continuing because "we already spent $X on this"?
  • Confirmation bias: Are you only looking for data that supports your preferred option?
  • FOMO: Are you doing this because "everyone else is"?
  • Availability bias: Are you overweighting a recent vivid event?

See: references/cognitive-debiasing.md

3. Competitive War Gaming

Don't just analyze yourself — predict the opponent:

  • If we cut prices, will competitors match or differentiate?
  • If we enter their market, will they counterattack or play defense?
  • If we acquire this company, how will competitors respond?

See: references/war-gaming.md

4. Stakeholder Playbook

Executive decisions aren't just about "right vs wrong" — they're about "who supports vs who opposes":

  • Board: Who is your ally? Who opposes? Who is the swing vote?
  • Investors: What metrics do they care about? What's their tolerance?
  • Leadership team: Who benefits / loses from this decision?
  • Key employees: Who might leave because of this?

See: references/stakeholder-playbook.md

5. Quantitative Analysis Tools

Not just qualitative — quantitative support built in:

  • Monte Carlo simulation: Probability distributions under uncertainty
  • ICE scoring: Impact / Confidence / Ease for quick prioritization
  • NPV / IRR: Investment return analysis
  • Expected Value: Probability-weighted outcome calculation
  • Scenario Planning: Best / base / worst case analysis

See: scripts/analysis_tools.py

Use Cases

Strategic Decisions

Resource Allocation

Crisis Response

  • Key executive departure, production outage, PR crisis — what to do first?
  • Example: CTO Departure Crisis

Competitive War Gaming

  • If we do X, how will competitors respond?

More examples: examples/

Quick Start

Just Ask

Treat it as your CEO chief of staff:

"Should we enter the Japan market?"
"5 projects, budget for 2 — how to prioritize?"
"CTO is leaving for a competitor. What do I do in the next 2 hours?"
"If we cut prices 20%, how will competitors respond?"

What to Expect

Not generic advice, but:

  1. Decision classification — What type of decision is this? Why does it matter?
  2. Bias check — What are your blind spots?
  3. Option comparison — At least 3 options + do nothing, with trade-offs
  4. Competitive / stakeholder analysis — How will others react?
  5. Recommendation + Next Steps — What to do, and what comes next?

Installation

OpenClaw / Claude Code / Cursor

Load SKILL.md directly:

# OpenClaw
openclaw skill install https://github.com/AIPMAndy/CEOskill

# Claude Code / Cursor
Add SKILL.md to your project context

Other Agent Runtimes

See: RUNTIME.md

Core principle:

  • Text-only runtimes can run the basic workflow
  • Web search makes strategic output stronger
  • Python / code execution unlocks quantitative analysis

Repository Structure

.
├── SKILL.md                            # Main skill definition (core)
├── references/
│   ├── frameworks.md                   # Decision framework library
│   ├── cognitive-debiasing.md          # Cognitive bias detection
│   ├── stakeholder-playbook.md         # Stakeholder management
│   └── war-gaming.md                   # Competitive simulation
├── scripts/analysis_tools.py           # Quantitative analysis tools
├── examples/                           # Real-world cases
│   ├── mna-acquisition-decision.md
│   ├── cto-departure-crisis.md
│   └── prioritization-five-initiatives.md
└── evals/evals.json                    # Evaluation cases

Roadmap

  • Main skill definition
  • Decision framework references
  • Cognitive debiasing
  • Stakeholder playbook
  • War gaming
  • Analysis tools
  • Add 10+ real-world cases (M&A / layoffs / fundraising / reorg)
  • Add benchmark / eval results
  • Add video demos
  • Add more Agent Runtime adapters

Contributing

Contributions welcome:

  • Real decision cases (anonymized)
  • New decision frameworks
  • Stronger eval cases
  • Agent Runtime adapters

Please read CONTRIBUTING.md first.

License

MIT License.

If This Helps You

  1. Give the repo a ⭐ Star
  2. Submit a real decision case or PR
  3. Share with someone who needs it

Turn AI decision support from a personal tool into a reusable decision toolkit.


Made with 🧠 by Andy