← ClaudeAtlas

ideationlisted

Generate ideas with structure when you're stumped — on what to build next, what the real problem is, or how to solve it. Pulls in research for context, diverges wide using matched frameworks, then converges on a prioritized few. Use when the user says "I'm stuck", "I'm stumped", "ideate", "brainstorm ideas", "help me think of", "what could I add", "what should I build", "I don't know what the problem is", "how could I solve", or wants idea generation on any topic (product, technical, business, writing, personal). For auditing an existing product against its users, use review-product instead.
nielsmadan/skills · ★ 0 · Code & Development · score 75
Install: claude install-skill nielsmadan/skills
<!-- Generated from https://github.com/nielsmadan/agentic-coding — edits here are overwritten. --> # Ideation Help someone who is stumped generate ideas — smartly and with structure, not a flat list of the obvious. Works on **any** topic: features for an existing product, what problem to even solve, how to solve a defined problem, business moves, writing, personal decisions. The method is always the same rhythm: **frame the stuck-state → gather just enough context (research if needed) → diverge wide → push past the obvious → converge on a prioritized few.** The frameworks change with the stuck-state; the rhythm doesn't. This is **divergent-then-convergent**: generate broadly with judgment off, *then* narrow. It is the broad counterpart to `review-product` (which audits a product against its users in a specific direction). Where `superpowers:brainstorming` gates implementation behind an approved design, this skill is upstream of that — it generates the ideas you'd then design. `review-product` delegates here when its audit surfaces a **wide-open gap** — a user job the product doesn't serve at all, whose fix isn't obvious — and needs a divergent spread of approaches rather than a single evidence-anchored fix. It passes the specific job as the topic and signals that the job is an already-validated unmet need, so the router lands in solution mode and skips re-litigating whether the job is worth solving. That hand-off is the main programmatic caller; a single named job must w