Design panel
Generate N independent design approaches, score them with a judge panel, synthesize the winner - for architecture or design decisions where the solution space is wide and a wrong pick is expensive (system design, data model, API shape, major refactor strategy). Pass the design brief as args.From its SKILL.md
npx -y skills add travelhawk/forgeshift --skill design-panelAssembled from the repository path, not quoted from the project. Check it against their README if it does not work.
One thing to look at
- 2 stars2 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.
SKILL.md
1.6 KB, as published. Nobody here has run it
/forge:design-panel — independent designs → judged → synthesized
Fire the design-panel workflow: independent designers with different priors beat one
design iterated (iteration anchors on the first idea). Default is the lean panel
(3 designers + 1 judge-synthesizer); panel: "wide" runs 4 designers + 3 voting judges +
a separate synthesis — reserve it for the most expensive decisions.
- Resolve the plugin home once:
FORGE_HOME="${CLAUDE_PLUGIN_ROOT:-$(forge-home)}". Pass the target repo asdir(absolute) — workflows do NOT follow the shellcd. - Parse
$ARGUMENTSinto the brief (problem, constraints, context), or an object{dir, brief, panel}. If the brief may not arrive intact, write it todesign-panel.input.mdin the product root first — the workflow reads it as a fallback; delete it after. - Invoke: Workflow tool,
scriptPath: "$FORGE_HOME/.claude/workflows/design-panel.js",args: {dir: "<abs product root>", brief: "<parsed>", panel: "<lean|wide>"}. - Relay the synthesized recommendation with the runner-up trade-offs, so the choice and its cost are both visible.
Next → record the decision in an ADR, then /forge:feature (or /forge:build) to
implement it.