Decision Stage

Input: $ARGUMENTS

Step 1: Frame the Decision

State what’s actually being decided. Most decision problems are poorly framed.

DECISION FRAME:
- The decision: [state it as a clear question]
- What triggered this: [why a decision is needed now]
- Who decides: [who has authority to commit]
- Reversibility: [easily reversible / partially reversible / irreversible]
- Timeline: [when must this be decided by]

Rule: If you can’t state the decision as a question, you’re not ready to decide. Go back to /anst.

Step 2: Enumerate Options

List all viable options. Don’t evaluate yet — just enumerate.

OPTIONS:
1. [Option A]: [brief description]
2. [Option B]: [brief description]
3. [Option C]: [brief description]
4. [Do nothing]: [what happens if no decision is made]

Rules:

• Always include “do nothing” as an explicit option
• Aim for 3-5 options. Fewer than 3 means you haven’t looked hard enough. More than 5 means some can be merged
• Include at least one option you’re emotionally resistant to

Step 3: Apply Decision Criteria

Define what matters and score each option.

CRITERIA:
1. [criterion 1]: weight [high/medium/low]
2. [criterion 2]: weight [high/medium/low]
3. [criterion 3]: weight [high/medium/low]
4. [criterion 4]: weight [high/medium/low]

EVALUATION:

| Option | [C1] | [C2] | [C3] | [C4] | Overall |
|--------|------|------|------|------|---------|
| A | [score] | [score] | [score] | [score] | [assessment] |
| B | [score] | [score] | [score] | [score] | [assessment] |
| C | [score] | [score] | [score] | [score] | [assessment] |
| Do nothing | [score] | [score] | [score] | [score] | [assessment] |

Use: strong / moderate / weak for scores. Avoid false precision with numbers.

Step 4: Check for Decision Traps

Actively look for biases distorting the evaluation.

DECISION TRAP CHECK:

| Trap | Risk | Detected? |
|------|------|-----------|
| **Sunk cost** | Favoring an option because of past investment | [yes/no] |
| **Status quo bias** | Favoring "do nothing" out of inertia | [yes/no] |
| **Framing effect** | The way options are described biases choice | [yes/no] |
| **Anchoring** | Over-weighting the first option considered | [yes/no] |
| **Confirmation bias** | Seeking evidence for a preferred option | [yes/no] |
| **Narrow framing** | Missing options by defining the choice too tightly | [yes/no] |

TRAPS DETECTED: [list any that apply]
CORRECTION: [how to adjust for detected traps]

SKIP: If this is a low-stakes, easily reversible decision, skip trap checking.

Step 5: Make the Call

Commit to a decision.

DECISION: [Option X]

RATIONALE:
- Primary reason: [the strongest argument for this option]
- Secondary reason: [supporting argument]
- What was sacrificed: [what this option gives up]
- What would change this: [conditions that would warrant revisiting]

Rule: State the decision clearly. Hedging (“maybe we should sort of lean toward…”) is not deciding.

Step 6: Document for Future Reference

Create a record that explains the decision to your future self.

DECISION RECORD:

Date: [today]
Decision: [one-line summary]
Context: [what was happening when this was decided]
Options considered: [list]
Chosen: [option] — because [core reason]
Rejected: [options] — because [core reasons]
Review trigger: [when or under what conditions to revisit]

Integration

Use with:

• /anst -> Complete analysis before entering the decision stage
• /exst -> Move from decision to execution
• /dcp -> For more rigorous decision process on high-stakes choices
• /cba -> For cost-benefit analysis as part of criteria evaluation
• /rfst -> Reflect on the decision after outcomes are known