brainstorming
brainstorming
You MUST use this before any creative work - creating features, building components, adding functionality, modifying behavior, or planning anything new, in software or out of it (a talk, a business, a renovation).
s4h-analogy-structure-mappingIdentifies the deep structural correspondence between two situations — genuine isomorphism vs superficial similarity. Triggers: 'does this analogy hold', 'map the structure', 'is this really like that', 'structural similarity', 'test the comparison'.
Independent GitHub snapshot · Automated classification · Original documentation governs
Analogies are persuasive but often wrong. The surface similarity that makes a comparison feel apt can obscure structural differences that make it invalid. This skill maps the deep structure of both situations and checks correspondence element by element — distinguishing genuine isomorphism (same structure, different content) from superficial similarity (same surface, different structure).
Step 1: State the Two Situations Write Situation A and Situation B clearly. State the analogy as claimed: "A is like B because..."
Framing check: Confirm the specific analogy before continuing. State what you've identified — the two situations being compared and the analogical claim linking them — in one sentence, then use AskUserQuestion:
Step 2: Extract Structure of Each For each situation independently, map: key actors, their relationships, the dynamics (what drives change), the constraints (what limits action), and the goals (what success looks like). Do this for each situation before comparing — comparison before extraction introduces bias.
Step 3: Map Elements A to B For each structural element in A, identify the corresponding element in B. State the mapping explicitly: "In A, X plays the role of Y in B."
Step 4: Classify Each Mapped Pair For each mapped pair: is the correspondence genuine (same structural role, same relationship type, same dynamics) or superficial (same label or surface feature, different structural role)?
Step 5: Find Where the Mapping Breaks Every analogy breaks somewhere — this is not a failure, it is where the analysis becomes useful. Which differences are structurally significant? Which make the analogy unreliable for the specific prediction or decision being made?
Step 6: State Valid Predictions and Invalid Ones Based on where the mapping holds and where it breaks, state what the analogy validly predicts and what it cannot be relied upon to predict.
Before proceeding, use the AskUserQuestion tool. State your interpretation of the situation in 1–2 sentences — what is being analyzed and what the core question is — then ask:
Proceed based on their selection. If the user reframes, incorporate the correction before running any analysis.
Situation A structure: [actors / relationships / dynamics / constraints / goals] Situation B structure: [actors / relationships / dynamics / constraints / goals]
Element mapping:
| Element in A | Corresponding element in B | Genuine / Superficial | Reason |
|---|---|---|---|
Where the mapping breaks:
[Specific structural differences and why they matter]
Valid predictions (analogy holds):
[What can be reliably inferred from B about A]
Invalid predictions (analogy breaks):
[What cannot be inferred — where relying on the analogy would mislead]
Genuine structural correspondence requires that the relationship between elements matches — not just that individual elements can be paired. Two things can share every node and differ entirely on the edges that connect them.
After delivering this output, use AskUserQuestion to offer the next move:
/s4h-logic-check — Validate that the structural mapping holds/s4h-analogy-domain-transfer — Use the structure map to guide a domain transfer/s4h-systems-feedback-mapping — Check if the structure reveals feedback dynamics