Agent Skills by yulong
Showing 22 vetted skills indexed across 3 GitHub repositories.
andrej-karpathy-perspective
Apply Karpathy's mental models to analyze AI topics.
steve-jobs-perspective
Analyze product strategy and critique decisions using Steve Jobs' thinking framework.
trump-perspective
Analyze political strategy using Trump-centered mental models and role-play rules.
zhangxuefeng-perspective
Analyze education and career decisions using Zhang Xuefeng's mental models.
mrbeast-perspective
Analyze YouTube content strategy and generate data-driven optimization plans.
zhang-yiming-perspective
Emulate Zhang Yiming's decision framework for structured strategic analysis.
taleb-perspective
Apply Taleb's framework to identify tail risks and asymmetric outcomes in strategic decisions.
ilya-sutskever-perspective
Analyze AI strategy and safety questions from Ilya Sutskever's perspective.
paul-graham-perspective
Apply Paul Graham's thinking framework to startup and writing questions.
feynman-perspective
Apply Feynman-inspired reasoning to critique claims and verify understanding.
munger-perspective
Classifies problems and researches evidence to produce Munger-style blunt insights and action steps.
naval-perspective
Classify problems, research Naval-style facts, and produce decisive conclusions.
elon-musk-perspective
Analyze costs and decisions using first principles reasoning and idiot index calculations.
clickhouse-io
Document ClickHouse table design patterns and query optimization techniques with SQL examples.
coding-standards
Standardize naming, architecture, and quality gates across TypeScript, JavaScript, React, and Node.js projects.
backend-patterns
Apply backend architecture patterns to design scalable Node.js APIs.
continuous-learning
Extract reusable patterns from Claude Code sessions into learned skills.
tdd-workflow
Enforce test-driven development across code changes with unit, integration, and E2E tests.
frontend-patterns
Provide reusable UI patterns for React and Next.js development.
security-review
Review authentication, input handling, secrets, and API design against security best practices.
strategic-compact
Prompt for manual context compaction at configurable thresholds via a PreToolUse hook.
code-understanding
Analyze codebases to generate hierarchical README.md documentation per directory.