IAW - In Another Way

Input: $ARGUMENTS

Core Principles

Intent and approach are different things. “I want to grow revenue” (intent) can be achieved by raising prices, selling more, entering new markets, or reducing churn (approaches). Users often fuse their intent with their first-thought approach. This skill separates them.
Synonyms are not alternatives. “Grow the team” rephrased as “expand headcount” is a synonym, not an alternative. A genuine alternative would be “automate instead of hiring” or “partner instead of building.” Alternatives must change the approach, not the wording.
The best alternative often reframes the problem. “How do I make meetings shorter?” → alternative: “How do I eliminate the need for this meeting?” The reframe dissolves the original constraint.
Alternatives have different consequences. Equivalent intent doesn’t mean equivalent side effects. Each alternative forecloses different options, costs different amounts, and risks different failures. Surface these.
Generating alternatives is easier than evaluating them. The skill must do both — generate genuinely different approaches AND rank them so the user can choose, not just browse.

Phase 1: Intent Extraction

[W1] ORIGINAL: [what the user said — the goal, claim, or plan, quoted]
[W2] SURFACE_APPROACH: [the approach embedded in the original statement]
[W3] UNDERLYING_INTENT: [what the user actually wants — separate from how]
[W4] INPUT_TYPE: [goal | claim | plan | strategy | method]

Separation Test

Ask: “If I could achieve the intent without this specific approach, would the user be satisfied?”

If YES → the approach is a means, not the goal. Generate alternatives to the approach.
If NO → the approach IS the goal. Generate alternative framings of the goal instead.

Phase 2: Alternative Generation

Generate genuinely different alternatives using multiple lenses:

[W-N] ALTERNATIVE: [description]
  LENS: [which generation method produced this]
  PRESERVES: [which aspects of the intent are preserved]
  CHANGES: [what's fundamentally different from the original]

Generation Lenses

Lens	Method	Example
Means substitution	Same goal, different method	”Grow revenue” → “Grow revenue by reducing churn instead of acquiring”
Inversion	Do the opposite of the original approach	”Add features” → “Remove features (simplify to increase value)“
Elimination	Remove the need for the goal entirely	”Faster meetings” → “Eliminate this meeting”
Level shift	Operate at a different level	”Teach this concept” → “Create an environment where they discover it”
Stakeholder shift	Different actor does it	”I’ll build it” → “Partner with someone who already built it”
Temporal shift	Different timing or sequence	”Launch then iterate” → “Iterate then launch”
Constraint change	Accept different tradeoffs	”Fast and cheap” → “Slow and excellent”
Decomposition	Achieve the intent in parts	”Big launch” → “Soft launch in one segment, then expand”

Phase 3: Alternative Assessment

For each alternative:

[W-N] ASSESSMENT: [alternative ref]
  INTENT_PRESERVED: [fully | mostly | partially]
  FEASIBILITY: [high | medium | low]
  CONSEQUENCES: [what's different about the side effects]
  FORECLOSES: [what options this alternative removes]
  ADVANTAGE: [what this alternative does BETTER than the original]
  DISADVANTAGE: [what this alternative does WORSE than the original]

Phase 4: Ranking

[W-N] RANKING (by fit to intent + feasibility):
  1. [alternative] — FIT: [level] — FEASIBILITY: [level] — KEY_ADVANTAGE: [what]
  2. [alternative] — FIT: [level] — FEASIBILITY: [level] — KEY_ADVANTAGE: [what]
  3. [alternative] — FIT: [level] — FEASIBILITY: [level] — KEY_ADVANTAGE: [what]
  ...
  ORIGINAL: [where does the original approach rank?] — [why]

Phase 5: Output

IN ANOTHER WAY
==============

ORIGINAL: [quoted]
INTENT: [underlying intent, separated from approach]

ALTERNATIVES (ranked by fit):
  1. [alternative]
     PRESERVES: [what stays the same]
     CHANGES: [what's different]
     ADVANTAGE: [why this might be better]
     DISADVANTAGE: [what's worse]

  2. [alternative]
     PRESERVES: [what stays the same]
     CHANGES: [what's different]
     ADVANTAGE: [advantage]
     DISADVANTAGE: [disadvantage]

  3. [alternative]
     ...

ORIGINAL RANK: [where the original approach falls and why]

REFRAME: [if a reframe exists that dissolves the original problem, state it here]

READY FOR:
- /cmp [top alternatives] — to compare the best alternatives rigorously
- /decide [alternatives] — to choose between them
- /ar [specific alternative] — to explore what follows if this alternative is right
- /aw [specific alternative] — to stress-test a promising alternative

Failure Modes

Failure	Signal	Fix
Synonym generation	”Alternatives” are just rewordings	Each alternative must change the APPROACH, not the wording
Intent lost	Alternatives don’t preserve the underlying goal	Check: does this alternative achieve what the user actually wants?
Only obvious alternatives	First-thought variants only	Use all generation lenses, especially inversion and elimination
No ranking	Flat list of alternatives without evaluation	Each alternative must be assessed and ranked
Consequences ignored	Alternatives listed without side-effect analysis	Every alternative has different consequences — surface them
Original not ranked	Alternatives listed but original not compared	The original might still be best — it should appear in the ranking
Reframe missed	No attempt to dissolve the problem	Always check: can the need for this be eliminated?

Depth Scaling

Depth	Min Alternatives	Generation Lenses	Assessment Depth	Ranking
1x	3	3	Fit only	Simple rank
2x	5	5	Fit + feasibility + consequences	Ranked with advantages
4x	8	7	Full assessment	Ranked with full comparison
8x	12	All 8	Full + second-order effects	Ranked with scenario analysis

Default: 2x. These are floors.

Pre-Completion Checklist

Intent separated from approach
Alternatives change the approach, not just the wording
Multiple generation lenses used (not just means substitution)
Reframe/elimination option explored
Each alternative assessed for consequences and tradeoffs
Alternatives ranked by fit and feasibility
Original approach included in ranking
Advantages AND disadvantages stated for each alternative

Integration

Use from: when current approach feels stuck, when exploring options
Routes to: /cmp, /decide, /ar, /aw for evaluating top alternatives
Complementary: /poa (possibility analysis — broader exploration)
Differs from /poa: poa explores what’s possible; iaw re-expresses a specific thing differently
Differs from /ma: ma generates from scratch; iaw generates alternatives to something specific
Differs from /va: va asks “what’s the opposite?”; iaw generates multiple alternatives, not just inversion

iaw - In Another Way