What problem does it solve? Decisions, plans, and claims often get generic answers when they need a specific kind of scrutiny. This Skill provides seven one-word reasoning lenses that force a particular mode of thinking, so a plan gets adversarial review, a claim gets evidence-checked, or a fork gets explicit tradeoffs instead of a vague reply. ## Core Features & Use Cases - Seven named lenses: DEVILS (argue against), RECEIPTS (demand evidence), ELI5 (plain explanation), PREMORTEM (failure autopsy), TRADEOFFS (pros/cons of options), SOCRATES (clarifying questions), FIRST (first-principles reasoning). - Two invocation modes: the owner types a lens name before a question, or the agent applies the right lens unprompted during planning, viability checks, and reviews. - Composable frames: lenses can be chained, such as running PREMORTEM on a plan and then DEVILS on the mitigation. - Use Case: Before shipping a launch, type "PREMORTEM: our new onboarding flow" to get a written autopsy of how it failed three months out, surfacing risks like distribution gaps before committing. ## Quick Start Type a lens name followed by your question, such as "DEVILS: should we build this feature in-house?", to get an answer framed entirely by that reasoning mode.
