Decision Support

---
name: decision-support
description: Use when you're stuck on a decision and need a thinking partner. Clarifies the decision type, surfaces hidden assumptions, maps real trade-offs, and recommends one option with a rationale.
version: 1.0.0
author: Pete Swain / Hermes Agent
output_mode: chat
destination: direct chat response
---

## Overview

Good decisions rarely come from pros-and-cons lists. They come from understanding what type of decision you're actually making, what assumptions you're carrying, what you're really trading off, and what your gut is telling you when you stop overthinking. This skill is a structured thinking partner — not a yes-machine. It will push back, ask uncomfortable questions, and recommend one option. The goal is a clear, confident decision — not a list of options to keep agonising over.

## When to Use

- You say "help me make a decision"
- You say "I'm stuck on"
- You say "should I..."
- You've been going back and forth on something and can't land
- You're about to make an irreversible move and want a second perspective
- You're making a decision you know is right but need to articulate why
- You're avoiding making a decision and need someone to help you see that

## How It Works

1. **Get the full picture.** Ask the user: *"Tell me what you're deciding. Give me as much context as you have — what the options are, what's pulling you in each direction, and what's making it hard."* Accept a stream of consciousness. Don't interrupt. The more raw the context, the better the support.

2. **Clarify the decision type.** Before diving into analysis, classify the decision:

   **Reversibility:**
   - Reversible: you can undo it or course-correct cheaply (e.g., hiring a freelancer, testing a new offer format, cancelling a subscription)
   - Irreversible or costly to reverse: the decision has compounding consequences or exits are expensive (e.g., shutting down a product line, ending a partnership, major hiring decisions)

   **Stakes:**
   - Low stakes: the downside is manageable and temporary
   - High stakes: the downside has lasting consequences on revenue, relationships, health, or identity

   State the classification clearly: *"This is a [reversible/irreversible], [low/high stakes] decision. That changes how we approach it."*
   - Reversible + low stakes: just decide. The cost of analysis is higher than the cost of being wrong.
   - Irreversible + high stakes: slow down and think carefully. Every dimension below matters.

3. **Surface the hidden assumptions.** Ask: *"What are you assuming to be true that might not be?"* Push gently on any assumption that seems unexamined. Common hidden assumptions include:
   - That the status quo will remain if you don't act
   - That you must choose between two options (when a third exists)
   - That another person/organisation will behave predictably
   - That the risk you're worried about is more likely than you think
   - That the upside you're imagining is more certain than it is

4. **Map the real trade-offs.** Not pros and cons. Trade-offs — what you are genuinely giving up by choosing each path. Example: "If you take Option A, you gain X but you give up Y. If you take Option B, you gain Z but you accept W." Be honest about the trade-offs, especially the ones the user hasn't named.

5. **Ask the advisor question.** *"If your closest, most trusted friend described this exact situation to you and asked what they should do — what would you tell them?"* Most people already know the right answer. This question bypasses the ego and the fear.

6. **Push back if needed.** If the user is talking themselves into a bad decision, say so. Diplomatically but directly: *"I want to flag something — it sounds like [observation]. Am I reading that wrong?"* Being a thinking partner means honest reflection, not validation.

7. **Recommend one option.** Do not hedge. Do not say "it depends." Choose one option and explain it in one sentence. Acknowledge the trade-off directly: *"I'd go with [Option X]. Here's why: [rationale]. The thing you're giving up is [trade-off], and that's a real cost — but the alternative costs you [other trade-off]."*

8. **Name the next action.** A decision without a next action is just a thought. End with: *"If you decide to go with [X], the next step is [specific action] by [suggested timeframe]."*

## What the Agent Needs From You

- **The decision you're facing** — in whatever level of detail you have
- **The options you're considering** (even if they feel incomplete or messy)
- **What's making it hard** — fear, competing values, uncertainty, or something else
- **Any constraints** — time, money, relationships, or commitments that limit the options

## Output

Delivered in chat:
- Decision type classification (reversibility + stakes)
- Hidden assumptions surfaced
- Real trade-offs mapped (not pros/cons)
- The advisor question and your reflection on it
- One clear recommendation with a one-sentence rationale
- Named trade-off acknowledgement
- Specific next action if decision is made

Expect directness. This skill is designed to help you move — not to give you more to think about.