ai-agent-skills

Use this skill when the user asks for analytical thinking (including naming it or directing use/apply/run with obvious misspellings; decisive) or wants a structured breakdown—decomposing the problem, defining metrics and hypotheses, organizing evidence, and synthesizing findings with explicit uncertainty. Use for quant-style reasoning framing, root-cause trees, decision tables, or comparable structure, including informal or incomplete data asks. Skip when they want open-ended idea spray with no measurement or hypothesis angle, or a short verdict with no decomposition requested.

Use this skill when the user asks for creative thinking (including naming it or directing use/apply/run with obvious misspellings; decisive) or wants divergent ideation—fluency, flexible perspectives, novel combinations, and elaboration, with optional light convergence. Use when they want fresh ideas, blue-sky options, reframes, or more variety before committing, including casual or messy prompts. Skip when they want a single delivered answer with no exploration, audit-only teardown with no generation asked for, or purely mechanical execution.

Use this skill when the user asks for critical thinking (including naming it or directing use/apply/run with obvious misspellings; decisive) or wants to evaluate a claim, argument, plan, or belief: clarify assertions, weigh evidence, surface assumptions, test reasoning for gaps or fallacies, scan biases, consider alternatives, and stress-test conclusions—whether they phrase it plainly ("red team", "devil's advocate", "what am I missing", steel/straw man, bias scan) or indirectly (decision-quality review, epistemic calibration). Skip for execution-only tasks with no evaluative angle, or when they only want wording, tone, layout, or open-ended brainstorming with no request to audit reasons, assumptions, or evidence.

Use this skill when the user asks for design thinking (including naming it or directing use/apply/run with obvious misspellings; decisive) or wants human-centered exploration—empathizing with needs, framing the problem, ideating, prototyping intent, and defining what to learn next. Use for HCD, service or UX concept sprints, how-might-we style discovery before build, or reframing from user evidence, even with messy context. Skip when the spec is fully frozen and they want no discovery, or when the task is code-only maintenance with no user problem framing requested.

Use this skill when the user asks for ethical thinking (including naming it or directing use/apply/run with obvious misspellings; decisive) or wants a structured pass on values and harms—mapping stakeholders, tradeoffs, power asymmetries, harms and benefits, consent, justice, and fair process for a plan or product. Use for moral review, fairness or AI-ethics style questions, stakeholder harm scans, or should-we questions beyond pure legality, including indirect asks. Skip when they want legal advice as such, only neutral facts with no normative review requested, or implementation-only work with no values lens asked for.

Use this skill when the user asks for first-principles thinking or first principles (including naming them or directing use/apply/run with obvious misspellings; decisive) or wants to reason from bedrock—stripping borrowed analogies and convention, surfacing fundamentals, then rebuilding the reasoning chain and implications. Use when they want to reason from scratch, challenge industry defaults, want physics-style business breakdowns, or sanity-check whether copying incumbents still makes sense, even if they never say first principles. Skip when they want a quick convention-following checklist with no rebuild of assumptions, or purely social coordination with no modeling ask.

Use this skill when the user asks for lateral thinking (including naming it or directing use/apply/run with obvious misspellings; decisive) or wants to break a sticky pattern—provocations, concept fans, challenge questions, and deliberate reframes to escape local optima. Use when ideation feels stuck, options feel incremental, or they want surprising angles (including after conventional brainstorming stalls), even if they never say "lateral thinking" or "provocation". Skip when they need a single compliant rule lookup, tight logical proof with no exploration phase, or no appetite for playful what-if moves.

Use this skill when the user asks for six thinking hats, parallel thinking, or Edward de Bono's method (including naming any of these or directing use/apply/run with obvious misspellings; decisive)—or wants a sequential, six-role thinking partner for one focus—keeping facts, feelings, risks, benefits, and new ideas in separate passes instead of one mixed reply. Use for structured brainstorms, decisions, or reviews when they ask for hat-by-hat analysis or name the method or author, even if phrasing is casual or incomplete. Skip for plain factual questions, execution-only implementation with no structured perspective pass, or one-off hot takes and single-angle verdicts where they do not want a fixed, hat-by-hat sequence across six roles.

Use this skill when the user asks for strategic thinking (including naming it or directing use/apply/run with obvious misspellings; decisive) or wants direction under constraint—connecting intent, context, capabilities, and options into coherent bets, tradeoffs, risks, and a sequenced path. Use when they talk about competitive positioning, roadmap narrative, where to play and how to win, portfolio prioritization, or leadership narrative that links goals to constraints, even if they never use the word strategy. Skip for step-by-step implementation detail when no strategic framing or choice among directions is requested.

Use this skill when the user asks for systems thinking (including naming it or directing use/apply/run with obvious misspellings; decisive) or wants to see how parts connect—mapping interdependencies, feedback loops, delays, stocks and flows, and leverage points before recommending action. Use when they worry about unintended consequences, holistic views, root causes beyond single blame, or how choices propagate across teams, products, or policies, even if they never say "feedback loop" or "stock and flow". Skip for single-step linear tasks, lone-variable calculations, or fixes that need no map of interactions or incentives.