Synthesize

Multiple brainstorm rounds produce a chronological stream of options. That's hard to compare. This skill reorganizes them by theme so the structure of the solution space becomes visible.

Why This Works

Brainstorming in rounds is powerful — later rounds displace earlier favorites, and cross-pollination between ideas creates options that neither round would produce alone. But the chronological output obscures patterns. Synthesizing by theme reveals:

Which categories have the most options (where the design space is richest)
Which categories have zero options (blind spots)
How thinking evolved across rounds
The true top contenders, ranked against the full field

On Activation

Mine — Collect every option from the conversation.
Categorize — Group by emergent theme, not by round.
Tabulate — Compact table per category.
Crown — Winners table with clear ranking.
Reflect — Note how thinking evolved across rounds.

Phase 1: Mine

Scan the conversation for all brainstormed options. Summarize each one compactly.

Include options from ALL sources — /brainstorm invocations, informal suggestions, user-injected ideas, variations that emerged in discussion. Don't filter yet.

If the conversation contains only one brainstorm round, that's fine — synthesis still adds value by reorganizing and re-evaluating.

Phase 2: Categorize

Group options into themes that emerge from the options themselves. Do not use predetermined categories. Look for shared mechanisms, shared philosophies, shared scope, or shared trade-off axes.

Name each category with a verb phrase that captures the shared intent — not a mechanical label.

Phase 3: Tabulate

For each category, produce a compact table. Be opinionated in your assessments.

Order categories from strongest to weakest (category containing the overall winner goes first).

Phase 4: Crown

Present a winners table. Note complementary options that strengthen the winner rather than competing with it.

Explain why #1 beats #2.

Phase 5: Reflect

Note how thinking evolved across rounds. Did later rounds displace earlier favorites? Did user-injected context shift the design space? Were there surprising convergences?

This validates that multiple rounds were worth the investment.

Anti-patterns

Chronological listing — Organizing by round is concatenation, not synthesis. Organize by theme.
Dropping options — Include everything, even bad ideas. Seeing why they're bad is part of the landscape.
Generic categories — "Good ideas" and "Bad ideas" is not categorization. Find the structural axis.
Burying the winners — The winners table should be immediately scannable.
Inventing options — Synthesis organizes what exists. Don't sneak in new ideas (that's brainstorming).

The user has brainstormed. Synthesize now. Start with Phase 1: Mine the conversation for all options.

synthesize