探索我们完整的 Claude 技能集合,扩展 AI 代理的能力。
Generates and edits high-quality images using the Google Gemini API with customizable aspect ratios, resolutions, and model preferences.
Replicates and resolves difficult bugs through a structured, iterative process of external testing and extensive logging.
Refines rough ideas into fully-formed software designs using structured questioning, alternative exploration, and incremental validation.
Builds and debugs web application features through a structured, iterative testing process using Playwright.
Builds new, persistent custom skills for coding agents by guiding the user through requirements, directory structure, and optional script bundling.
Extracts subtitles and transcripts from YouTube videos directly into local text files using command-line tools or browser automation.
Automates complex, long-running projects by decomposing them into sub-tasks and executing them across multiple self-continuing sessions.
Enables non-interactive, headless execution of programming and system tasks using the Claude Code CLI for full automation.
Evaluates agent skills against Anthropic's official best practices to ensure optimal structure, naming, and content organization.
Audits Claude Code skills against official best practices to ensure high-quality, structured, and efficient performance.
Standardizes and validates skill progression using international educational frameworks like CEFR and Bloom's Taxonomy.
Generates production-grade, pedagogically sound code examples using a reasoning-first approach to accelerate developer learning.
Audits and refines technical documentation to ensure maximum learner comprehension and accessibility across all proficiency levels.
Transforms session-specific fixes and insights into permanent organizational knowledge by updating core documentation and agent protocols.
Automates mandatory checks and fixes for R packages to ensure compliance with CRAN's strict ad-hoc submission requirements.
Facilitates the development of specialized Claude Code skills by providing structured workflows, design patterns, and optimization strategies for extending AI capabilities.
Creates distinctive, production-grade frontend interfaces with high design quality while avoiding generic AI-generated aesthetics.
Evaluates the quality of educational lessons and book chapters using a systematic weighted rubric to ensure technical accuracy and pedagogical effectiveness.
Generates pedagogically-aligned, proficiency-calibrated presentation decks from educational content using structured prompt engineering for NotebookLM.
Automates the configuration, local validation, and deployment of Docusaurus documentation sites to GitHub Pages using automated CI/CD workflows.
Identifies non-deterministic test failures by analyzing CI history and code patterns to improve software reliability.
Creates hybrid AI personas by merging multiple voice profiles using weighted interpolation and vocabulary reconciliation.
Empowers engineering teams to make data-driven technical choices through weighted decision matrices and automated Architecture Decision Record (ADR) generation.
Manages artifact versioning, ownership tracking, and lifecycle metadata for autonomous development workflows.
Validates software development lifecycle phase gates using multi-agent reviews and automated readiness reporting.
Analyzes GitHub repository structure, documentation quality, and dependency health to provide comprehensive codebase insights.
Generates executive-ready performance summaries and strategic insights from marketing and KPI data.
Generates tailored writing voice profiles from natural language descriptions to standardize AI output across different domains.
Manages system architecture changes through automated impact analysis, ADR generation, and phased migration planning.
Manages bounded autonomous coding loops by tracking retries, synthesizing error feedback, and escalating complex failures to developers.
Scroll for more results...