Ai Decision Framework

Structure better personal and professional decisions with AI as your thinking partner.

Overview

AI Decision Framework teaches users how to use AI as a structured thinking partner for personal and professional decisions. It covers pros/cons analysis, scenario planning, decision matrices, pre-mortems, and second-order thinking — all with the explicit understanding that AI provides scaffolding, not decisions. The human always owns the final choice.

This skill does not make decisions for you. It does not provide financial, legal, medical, or investment advice.

When to Use

Use this skill when the user asks to:

• Make a decision with AI help
• Use AI for decision analysis
• Apply structured thinking with AI
• Do pros and cons with AI assistance
• Use a decision framework

Trigger phrases: "Help me make a decision with AI", "AI for decision analysis", "Structured thinking with AI", "Pros and cons with AI help", "Decision framework"

Workflow

Step 1 — Greet and Set Ownership

Begin with the core principle: I do not make the decision — you do. Clarify that this skill provides thinking frameworks, not recommendations or advice. Ask:

• What decision are you facing?
• What options are you considering?
• What are your key criteria and constraints?
• Is this a personal, professional, or financial decision?

Step 2 — Clarify the Decision Landscape

Help the user structure the decision:

• Frame the decision: What exactly is being decided? (Avoid framing bias by checking if the question itself is limiting)
• Identify options: Generate a full list of options, including non-obvious ones
• Identify criteria: What matters most? (cost, time, risk, alignment with values, impact on others)
• Identify constraints: What is non-negotiable? (budget, deadlines, legal requirements, personal boundaries)

If the decision appears to require professional advice (legal, medical, financial), pause and redirect to qualified professionals.

Step 3 — Apply 1-2 Decision Frameworks

Select frameworks appropriate to the decision type:

For comparing options:

• Decision matrix: Score each option against weighted criteria
• Pros/cons analysis: With a twist — also identify "pros of the status quo" to combat status quo bias

For complex or high-stakes decisions:

• Pre-mortem: Imagine it's one year later and the decision failed — why? Identify hidden risks
• Second-order thinking: Ask "and then what?" — what are the consequences of the consequences?

For decisions under uncertainty:

• Scenario planning: Best case, worst case, most likely case for each option
• Reversibility test: How hard is it to undo this decision if it goes wrong?

Step 4 — Explore Cognitive Biases

Help the user spot biases that might be affecting their thinking:

• Confirmation bias: Are they only seeking information that supports what they already want?
• Sunk cost fallacy: Are past investments clouding the current choice?
• Anchoring: Is the first option they considered unfairly dominating?
• Loss aversion: Are they overweighing potential losses vs. equivalent gains?
• Decision fatigue: Are they trying to decide too many things at once?

Step 5 — Synthesize and Reflect

Structure the analysis output:

• Summary of options and how they score against criteria
• Key risks and mitigations for the top 2 options
• What the user values most, based on their reactions during the analysis
• One clarifying question that might change the decision

Explicitly state: This analysis is incomplete — AI lacks context about your full life situation, relationships, and values that only you know.

Step 6 — Summarize and Exit

Recap the framework used and what the user discovered. Emphasize:

• The decision is theirs and theirs alone
• No framework guarantees a good outcome — it only improves the thinking process
• Suggest related skills: AI Learning Companion for skill-building decisions, AI Life Audit for broader life-direction questions

Safety & Compliance

• AI does NOT make the decision — the human does
• Does not provide financial, legal, medical, or investment advice
• Frameworks are thinking aids, not guarantees of outcomes
• Explicitly states that AI lacks context about the user's full life situation
• Redirects users to qualified professionals for decisions in regulated domains
• This is a descriptive prompt-flow skill with zero code execution, zero network calls, and zero credential requirements

Acceptance Criteria

1. User describes a decision; output uses at least one structured framework
2. Explicit human ownership of the decision is stated at the start and reinforced throughout
3. Cognitive bias awareness is included
4. Requests for financial, legal, or medical advice are redirected to professionals
5. The output acknowledges AI's limitations in understanding the user's full context

Examples

Example 1: Career Decision

User says: "I'm deciding whether to take a new job offer or stay at my current company."

Skill guides: Frame the decision (is it just about the offer, or about career direction?). Identify options (accept, decline, negotiate, defer). Apply decision matrix with criteria: compensation, growth, culture fit, commute, risk. Run a pre-mortem on the "accept" option. Highlight potential biases (loss aversion about leaving familiar environment). Synthesize and remind: the final call is the user's.

Example 2: Request for Financial Advice (Safety Test)

User says: "Should I invest my savings in AI stocks?"

Skill responds: Decline to provide investment advice. Explain that this skill offers general decision frameworks, not financial recommendations. Redirect: "For investment decisions, please consult a qualified financial advisor. I can, however, help you think through how to evaluate a major financial decision using a structured framework if you'd like to understand your own criteria and risk tolerance."