Goal Reframing Procedure

Input: $ARGUMENTS

Overview

Transform “impossible” or problematic goals into achievable versions.

CORE PHILOSOPHY: There is no such thing as an impossible goal. There is only insufficient decomposition or incorrect framing.

“Impossible” usually means one of:

• We haven’t thought deeply enough about what’s actually required
• We’re using the wrong frame (literal vs metaphorical)
• We haven’t found the achievable version that captures the intent
• We said airplanes would take a million years in 1900

“Unethical” goals can almost always be achieved ethically:

• The underlying need is usually legitimate
• Only the METHOD was problematic
• Reframe to achieve the same outcome through ethical means

CRITICAL: Reframing vs Substitution

Reframing IS a form of goal change. This skill explicitly handles substitution.

Step 0: Capture Original and Get Consent

ORIGINAL GOAL (verbatim): "[user's exact words]"

Acknowledge upfront: “I’m going to propose a REFRAME of your goal. This means suggesting a different way to think about what you want. The reframed version may be a different goal that achieves what matters to you.”

Reframing Types

Consent Template for Reframes

REFRAME PROPOSAL
================
Your original goal: "[original]"

I'm proposing a reframe: "[reframed goal]"

This is a [SCOPE/INTENT/VALUE/PIVOT] reframe:
- What's preserved: [what stays the same]
- What changes: [what's different]
- Why I think it helps: [reasoning]

If you achieved the reframed goal but NOT the original, would that be:
[ ] Acceptable (the reframe captures what matters)
[ ] Not acceptable (I still want the original)
[ ] I need to think about it

Your choice?

When to Use

• Goal seems “impossible”
• Goal seems unethical or harmful
• Goal is too vague to act on
• Goal is expressed as outcome beyond control
• System is tempted to reject or give up

Context-Adaptive Variants

Reframe-Lite (URGENT)

Steps:

1. Capture original
2. One quick reframe attempt
3. If blocked, acknowledge and move on

Reframe-Standard

Steps: Full procedure with consent check

Reframe-Full (HIGH stakes)

Steps: Full procedure + multiple reframe options + user selects

Steps

Step 1: Understand Why Goal Seems Impossible

• What specific obstacle makes it “impossible”?
• Is it physically impossible, or just very hard?
• Is it impossible as stated, or as interpreted?

Step 2: Identify the Underlying Value

• What would achieving this goal give you?
• What’s the deeper need being served?
• → INVOKE: /ve if unclear

Step 3: Generate Reframe Options

Generate multiple reframes at different levels:

1. Scope Preservation Reframe: Same goal, remove unnecessary constraints
2. Intent Preservation Reframe: Different goal, same underlying value
3. Value Discovery Reframe: Different goal based on what we learned matters
4. Literal vs Metaphorical: Is the goal meant literally or as an expression?

Step 4: Present Options with Honest Assessment

For each reframe:

• What it achieves
• What it doesn’t achieve
• How it relates to original

Step 5: Get Explicit Choice

User selects:

• One of the reframes
• Original goal anyway (with acknowledgment of difficulty)
• Something else (iterate)

Verification

• Original goal captured verbatim
• Reframe type identified
• User consent obtained for non-trivial reframes
• User understands what’s preserved and what’s lost
• If honest rejection, reason is factual and specific

Niche Documentation

Where This Skill Works Best

• Goals that seem impossible but have achievable variants
• Goals where underlying value is different from stated goal
• Situations where user is stuck on one framing

Where This Skill Struggles

• Goals that are genuinely impossible (use honest rejection)
• Goals where user doesn’t want alternatives
• Time-critical situations (use Reframe-Lite)

Integration

• Often invoked from: /gu, /grf
• Routes to: /gd, /stg
• Gate: clarification_vs_substitution (but this skill is explicitly about substitution)