Compares 2+ candidate variants under parity discipline and returns a Pareto front (not a scalar winner) — declares the selection policy and parity plan BEFORE scoring, then scores each dimension across all variants in parallel to prevent anchoring bias. Make sure to use this skill whenever the user asks "A or B", "which is better", "compare X and Y", "trade-off between X and Y", "should we pick X or Y", "Pareto for these options", "PostgreSQL vs MySQL", "NATS vs Kafka", "library A vs library B" — anywhere two or more viable approaches are on the table and a fair, recorded comparison is wanted before committing. Also use when /h-explore has just produced a SolutionPortfolio. NOT for generating new variants (use h-explore first). NOT for committing to the winner — that requires manual /h-decide per Transformer Mandate.
$npx -y skills add m0n0x41d/haft --skill h-compareInstalls into the current project.
Run `npx skills use "https://github.com/m0n0x41d/haft" --skill "m0n0x41d/haft/h-compare"` and follow the generated skill instructions now. Read its complete output, redirecting it to a temporary file first if necessary. Resolve relative paths from the supporting-files directory it provides.
Use the skills in "https://github.com/m0n0x41d/haft" that are relevant to the current task. Run `npx skills add "https://github.com/m0n0x41d/haft"` and select the relevant skills, then follow their instructions.