Claude Code Skills Collection

Claude Code Plugin Available — Install all 74 skills instantly with /plugin marketplace add lyndonkl/claude then /plugin install thinking-frameworks-skills

A comprehensive collection of production-ready skills for Claude Code, covering thinking frameworks, decision-making tools, research methods, design patterns, and specialized domains.

Overview

This repository contains 74 skills designed to enhance Claude Code's capabilities across strategic thinking, product development, research, experimentation, and creative problem-solving. Each skill includes:

• Structured workflows with step-by-step guidance
• Practical templates for immediate use
• Advanced methodologies for complex scenarios
• Evaluation rubrics with quality criteria and common failure modes
• Progressive disclosure (SKILL.md → template.md → methodology.md)

Table of Contents

• Skills by Category
  • 🧠 Strategic Thinking & Decision Making
  • 💡 Problem Solving & Analysis
  • 🔬 Research & Discovery
  • 🗣️ Dialogue & Deliberation
  • 💭 Ideation & Creativity
  • 📊 Data & Modeling
  • 🔷 Geometric Deep Learning
  • 🏗️ Architecture & Design
  • 🔒 Security & Risk
  • 📝 Communication & Documentation
  • 🔬 Scientific Writing
  • 🎯 Estimation & Forecasting
  • 💼 Business & Product Management
  • ⏱️ Productivity & Learning
  • 🏢 Organizational Design
  • 🛡️ Ethics & Evaluation
  • 🍳 Specialized Domains
  • 🛠️ Skill Development & Meta-Tools
• Installation
• Using Skills
• Skill Development Status
• Key Features
• Contributing
• Resources
• License
• Acknowledgments

Skills by Category

🧠 Strategic Thinking & Decision Making

decision-matrix - Multi-criteria decision analysis with weighted scoring, sensitivity analysis, and uncertainty handling for complex choices between 3-10 options.

bayesian-reasoning-calibration - Probabilistic reasoning with prior/posterior updates, Bayes theorem application, and calibration techniques for reducing overconfidence.

alignment-values-north-star - Strategic alignment framework connecting daily decisions to core values and long-term vision using north star metrics and principle mapping.

environmental-scanning-foresight - Anticipate change through PESTLE analysis, weak signal detection, scenario planning (2x2 matrix), cross-impact analysis, and signposts for early warning systems.

expected-value - Rational decision-making under uncertainty through probability-weighted outcomes (EV = Σ p×v). Covers scenario identification, probability estimation (base rates, triangulation, calibration), payoff quantification (NPV, non-monetary factors), decision trees (fold-back induction, optionality), expected utility for risk aversion, sensitivity analysis, and bias mitigation (overconfidence, sunk costs, tail risk neglect).

heuristics-and-checklists - Practical decision-making through mental shortcuts (heuristics) and systematic error prevention (checklists). Design fast & frugal heuristics (recognition, take-the-best, satisficing), create effective checklists (5-9 killer items, READ-DO vs DO-CONFIRM formats), validate rules empirically (≥80% accuracy target), and mitigate cognitive biases (availability, representativeness, anchoring). Covers aviation, surgical, and software deployment checklists with proven 60-80% error reduction.

hypotheticals-counterfactuals - Explore alternative scenarios and test assumptions through "what if" thinking. Use counterfactual reasoning (backward-looking: "what would have happened if?") to understand causality and learn from decisions. Apply scenario planning (forward-looking: optimistic/baseline/pessimistic futures, 2×2 matrices) to prepare for uncertainty. Conduct rigorous pre-mortems (6-step process: imagine failure, identify causes, vote on risks, assign mitigations). Extract common actions, hedges, and options from scenarios. Define leading indicators with clear thresholds for monitoring which future unfolds. Covers minimal rewrite principle, causal mechanism specification, probability calibration, and stress testing decisions.

kill-criteria-exit-ramps - Define objective stopping rules for projects to avoid sunk cost fallacy and optimize resource allocation. Set upfront kill criteria (quantifiable metrics, time horizons, decision authority) before emotional/financial investment. Design go/no-go gates for milestone-based evaluation with increasing investment stages. Apply pivot vs. kill framework (customer pain validation, learning rate, burn rate sustainability, opportunity cost analysis). Manage project portfolios with quarterly ranking (EV/Cost ratio), systematic rebalancing, and bottom 20-30% kill threshold. Execute disciplined wind-downs (1 month max) with team reallocation, customer transition, and blameless postmortems. Use pre-mortem inversion ("would we start this today?") to overcome sunk cost bias. Normalize killing projects as capital allocation discipline. Covers behavioral economics of stopping, real options analysis, Bayesian updating for kill probability, and organizational culture change.

portfolio-roadmapping-bets - Strategic portfolio roadmapping across time horizons (H1/H2/H3: Now/Next/Later) with disciplined betting framework. Size bets by effort (S/M/L/XL) and impact (1x/3x/10x), sequence across horizons with dependency mapping, set exit/scale criteria for kill or double-down decisions, and balance portfolio using 70-20-10 rule (core/adjacent/transformational). Apply McKinsey Three Horizons, RICE/ICE scoring, critical path analysis, and staged funding models. Validate capacity feasibility (effort ≤ 80% capacity), ensure impact ladders to strategic theme with risk adjustment, and establish review cadence (monthly H1, quarterly H2, semi-annual H3) with kill/pivot/persevere/scale framework. Covers horizon planning, bet sizing methodologies, portfolio balancing techniques, dependency sequencing, and portfolio health metrics tracking.

postmortem - Blameless postmortem analysis transforming failures into learning. Document timeline with specific timestamps (detection → investigation → resolution), quantify impact across dimensions (users, revenue, SLA, reputation), conduct root cause analysis using 5 Whys or fishbone diagrams to reach systemic issues (not individual blame), define SMART corrective actions (specific, measurable, assigned, realistic, time-bound) using hierarchy of controls (eliminate, substitute, engineering controls, administrative, training). Maintain blameless tone focusing on systems/processes, conduct within 48 hours while memory fresh, track action items to >90% completion, share broadly for organizational learning. Apply to production outages, security incidents, product failures, project misses, or near-misses. Covers blameless culture principles, root cause techniques, corrective action frameworks, incident response patterns, facilitation techniques, and learning metrics.

prioritization-effort-impact - Transform overwhelming backlogs into actionable priorities using 2x2 effort-impact matrix (Quick Wins, Big Bets, Fill-Ins, Time Sinks). Score items on effort (1-5: time, complexity, risk, dependencies) and impact (1-5: users, business value, strategic alignment, user pain) with diverse stakeholder input (engineering estimates effort, product/sales estimate impact). Plot on matrix to identify Quick Wins (high impact, low effort - do first) vs Time Sinks (low impact, high effort - avoid). Sequence roadmap: Quick Wins first (build momentum), Big Bets second (strategic impact), Fill-Ins opportunistically, reject Time Sinks with clear rationale. Apply calibration techniques (reference examples, silent voting, forced ranking), validate with data (usage analytics, user surveys, NPS drivers), and plan capacity with 20-30% buffer. Alternative frameworks include RICE (Reach × Impact × Confidence / Effort), MoSCoW (Must/Should/Could/Won't), ICE scoring, Kano model, and cost of delay. Covers advanced scoring methodologies, stakeholder alignment techniques, roadmap optimization, dependency mapping, and common pitfalls (optimism bias, HIPPO pressure, scope creep).

project-risk-register - Proactively identify, assess, prioritize, and monitor project risks using structured risk register. Score risks on probability×impact (1-5 each, 5×5 matrix = 1-25 risk score) with thresholds: Critical (≥20), High (12-19), Medium (6-11), Low (1-5). For each risk: identify root cause (not symptoms), assign named owner (not "team"), define mitigation (reduce probability before) AND contingency (reduce impact if occurs), set quantifiable triggers for contingency activation. Brainstorm risks across 6 categories (technical, schedule, resource, external, scope, organizational) using structured checklists. Monitor regularly (weekly for active projects, monthly for longer) tracking new/closed risks, score changes with rationale, risk burn-down (total exposure decreasing), and leading indicators. Integrate mitigation actions into project schedule, budget contingency reserves based on Expected Monetary Value (EMV), and use Monte Carlo simulation for