| name | fpf-variants |
| description | Explore the solution space by generating ≥3 genuinely distinct variants within the chosen method family. Think through alternatives using the portfolio template as scaffolding. |
| argument-hint | [problem-or-decision-context] |
What this skill IS
This is creative solution exploration, not box-checking. You are exploring the solution space — generating diverse alternatives, characterizing trade-offs, identifying stepping stones. The goal is NOT to produce 3 trivially different options — it's to map the landscape of possible approaches.
Work through the template as scaffolding. Generate variants BY filling the template, not in chat first.
Prerequisites
- PROB-* or ANOM-* exists with acceptance spec
- STRAT-* exists with method family bet (T4 required; T3 recommended if available. If missing at T4: invoke
/fpf-sota first)
- CHR-* exists with comparison indicators (if missing: invoke
/fpf-characterize first)
Generation guidance
Diversity strategies (pick at least one):
- Decomposition: Break the problem into sub-problems, solve each differently
- Analogical: Import patterns from different domains/stacks
- Constraint relaxation: What if we dropped constraint X? What becomes possible?
- Inversion: Solve the opposite problem. What if we optimized for a different indicator?
- Extreme points: What's the simplest possible? Most sophisticated? Most unusual?
Test for genuine distinctness: If two variants differ only in a parameter value or minor implementation detail, they are NOT distinct. Distinct means different trade-off profiles — different points on the Pareto front.
