Idea quality audit
Decide which of a batch of candidate ideas are worth committing to, by scoring them on explicit quality dimensions and then adversarially stress-testing the strongest few.
graph LR nthinkdecisionoptionreview["Decision Option Review"] nthinkredteamlight["Red Team Light"] nthinkdecisionoptionreview --> nthinkredteamlight
Each step passes a compressed artifact to the next. The numbered list below is the same chain in text.
The chain
Section titled “The chain”Job: decide which of a batch of candidate ideas are worth committing to, by scoring them on explicit quality dimensions and then stress-testing the strongest few.
Use when: you have a pile of candidate ideas (from a brainstorm, a backlog, or an ideation skill) and need to separate the genuinely strong from the merely fluent before investing in any of them. It pairs naturally after an ideation skill (think-brainwriting, think-scamper) that generated the batch.
think-decision-option-review(skills/think-decision-option-review/SKILL.md)- Treat the ideas as options; score them against weighted idea-quality criteria (for example novelty, feasibility, value, specificity); surface the tradeoffs and a ranked shortlist.
- Carry forward: the top few ideas, and where the scoring was soft.
think-red-team-light(skills/think-red-team-light/SKILL.md)- Adversarially attack the top-ranked ideas - the strongest objections, steelmanned - to find which survive scrutiny and which only sounded good.
- Carry forward: the survivors and their open weaknesses.
Composite artifact
Section titled “Composite artifact”An idea-quality audit: a scored, ranked shortlist of the batch, with the strongest few stress-tested, and a verdict on which ideas are worth committing to. The point is that commitment goes to ideas that survive scrutiny, not to the ones that read best on first pass.
Why a recipe, not a skill
Section titled “Why a recipe, not a skill”“Idea-quality audit” has no separable cognitive mechanism of its own: scoring a batch is think-decision-option-review (options become ideas), and pressure-testing is think-red-team-light. It ships as this thin chain rather than a near-twin skill (the 2026-06-03 catalog vetting reached this verdict). Use think-decision-option-review on its own for a one-off scored comparison; reach for this recipe when you specifically want score-then-stress-test over a batch of ideas.
Token discipline
Section titled “Token discipline”Carry only the top-ranked shortlist between steps (the ranked few plus their soft spots, then the red-team survivors), not the full scoring matrix for every idea.