TEXT

High-Stakes Decision Support System

Contributed by mmanisaligil

Improved by Laravel Company · 2026-09-07

You are an expert System Architect and UX Designer specializing in cognitive behavioral frameworks. Your task is to design the complete specification for a high-stakes decision support application named "Pivot."

Product Philosophy: Pivot is not a simple pros/cons generator. Its core value lies in guiding the user through a rigorous, structured analytical process to surface hidden assumptions, evaluate complex trade-offs, and foster objective decision-making regarding major life and business choices. The output must prioritize analytical depth and intellectual rigor over simple recommendation.

System Requirements & Architecture:

  1. Decision Intake Module: The system must begin by capturing comprehensive input from the user. This intake must include:

    • The specific decision being made (the choice set).
    • Stated constraints (time, financial, relational, obligatory).
    • Stated core values (the user must identify their top 3 guiding values).
    • The user's current leaning/preference.
    • The imposed deadline.

  2. Mandatory Assumption Discovery (LLM Integration):

    • After intake, the system must trigger an LLM API call to generate exactly 5 highly specific, probing clarifying questions.
    • These questions must be expertly crafted to surface unstated assumptions, potential hidden trade-offs, and psychological biases related to the user's specific scenario.
    • The user must answer all 5 questions before the analytical phase can proceed.

  3. Core Analytical Frameworks (Six Modules):
    The system must present six distinct analytical frames, each functioning as a separate, detailed analysis module (ideally presented in navigable tabs or sections). Each frame must be powered by a separate, carefully crafted LLM API prompt designed to execute the specific analytical lens:

    • (1) Expected Value: Probability-weighted outcomes for each option.
    • (2) Regret Minimization: An analysis focusing on which option minimizes potential regret at a future point (e.g., age 80).
    • (3) Values Coherence: An assessment of which option is most consistent with the user's stated top 3 values, supported by explicit evidence from the input.
    • (4) Reversibility Index: An assessment of how easily each option can be undone or reversed if the decision proves incorrect.
    • (5) Second-Order Effects: A projection of the consequences following each option over defined timelines (e.g., 6 months and 3 years).
    • (6) Advice to a Friend: An empathetic simulation: what objective, reasoned advice would be given to a trusted friend in this exact situation?

  4. Counter-Argument Module:

    • Following the six analytical frames, the system must include a dedicated section titled "Devil's Advocate Brief."
    • This brief must generate a maximally rigorous argument against the user's current leaning, exploring all potential risks and flaws, arguing as strongly as possible.

  5. Decision Record & Iteration:

    • A persistent "Decision Record" must store the initial intake, all six analytical outputs, the Devil's Advocate Brief, and the final decision chosen.
    • The system must include a mechanism for post-decision updating, requiring the user to input the actual outcome of the decision at 90 days and 1 year post-decision.

Technical Stack & Design Constraints:

  • Frontend: React.
  • Data Persistence: LocalStorage for session and record management.
  • AI Interaction: All LLM interactions must be modularized, with one distinct, highly specialized prompt for each of the six analytical frames.
  • Design Tone: The entire interface must adhere to a focused, serious, and analytical design. There must be absolutely no gamification, motivational language, or superficial encouragement. The system is a tool for navigating reality, not for feeling good about the choice.

Deliverable: Provide the complete, detailed architectural specification and the necessary prompt engineering guidance required to implement the Pivot system based on these requirements.

Original prompt (before our improvements)

Build a high-stakes decision support system called "Pivot" — a structured thinking tool for major life and business decisions. This is distinct from a simple pros/cons list. The value is in the structured analytical process, not the output document. Core features: - Decision intake: user describes the decision (what they're choosing between), their constraints (time, money, relationships, obligations), their stated values (top 3), their current leaning, and their deadline - Mandatory clarifying questions: [LLM API] generates 5 questions designed to surface hidden assumptions and unstated trade-offs in the user's specific decision. User must answer all 5 before proceeding. The quality of these questions is the quality of the product - Six analytical frames (each run as a separate API call, shown in tabs): (1) Expected value — probability-weighted outcomes under each option (2) Regret minimization — which option you're least likely to regret at age 80 (3) Values coherence — which option is most consistent with stated values, with specific evidence (4) Reversibility index — how easily each option can be undone if it's wrong (5) Second-order effects — what follows from each option in 6 months and 3 years (6) Advice to a friend — if a trusted friend described this exact situation, what would you tell them? - Devil's advocate brief: a separate analysis arguing as strongly as possible against the user's current leaning — shown after the 6 frames - Decision record: stored with all analysis and the final decision made. User updates with actual outcome at 90 days and 1 year Stack: React, [LLM API] with one carefully crafted prompt per analytical frame, localStorage. Focused, serious design — no gamification, no encouragement. This handles real decisions.