すべてのClaudeスキルを閲覧

Defense in Depth Validation

Implements a multi-layered testing and validation framework to catch bugs across syntax, logic, integration, and user workflows.

Subagent Driven Development

Optimizes complex coding tasks by delegating focused subtasks to fresh AI agents to prevent context pollution and improve accuracy.

Root Cause Tracing

Eliminates technical debt by identifying and fixing the origin of software bugs rather than patching superficial symptoms.

Commit Message Generator

Generates natural, human-like Git commit messages by analyzing staged code changes and repository history.

Human-Style GitHub Operations

Automates Git workflows with natural, human-like commit messages that blend seamlessly into any developer's repository history.

Backend API Development

Architects and implements robust Next.js API routes and singleton service layers with built-in validation, authentication, and caching.

Verified Task Completion

Enforces strict validation of all coding tasks through live command output before allowing any claims of success.

Workflow Verifier

Verifies AI-generated workflow outputs and artifacts using a hybrid approach of structural scripts and LLM-as-judge assessments.

Plan Marshall Documentation Plugin

Enforces consistency and reliability in AsciiDoc documentation workflows within the Plan Marshall orchestration layer.

Canvas LMS Integration

Connects Claude Code to the Canvas LMS API to manage courses, track assignments, and monitor academic progress.

Claude Plugin Validator

Validates Claude Code plugin structures, configurations, and security practices to ensure seamless integration and deployment.

Smart Init

Initializes the ClaudeShack ecosystem by analyzing your codebase and conversation history to establish a deep, autonomous project baseline.

Smithery & Cloudflare Deployment Manager

Streamlines the deployment and management of Smithery MCP servers and Cloudflare Workers for Canvas integrations.

When Stuck

Provides structured diagnostic techniques and recovery strategies to help Claude overcome development roadblocks and avoid repetitive failure loops.

Writing Plans

Generates granular, step-by-step implementation plans using TDD principles to streamline complex development workflows.

Executing Plans

Implements multi-step development plans through controlled batch execution and mandatory verification checkpoints.

Verification Before Completion

Ensures code quality by requiring fresh command-line evidence before claiming any task is finished.

Documentation Wizard

Maintains living documentation by synchronizing READMEs, API references, and architecture records with your codebase and design decisions.

Database Schema Management

Manages PostgreSQL database schemas and Prisma ORM migrations within the ClaudeCode Sentiment Monitor project.

Claude Agent Skills

Provides comprehensive guidance and implementation patterns for creating modular, high-performance capabilities that extend Claude's functionality.

Executing Plan System

Implements complex technical plans through structured batch execution and mandatory review checkpoints to ensure implementation accuracy.

Documentation Wizard

Maintains, updates, and generates accurate project documentation by cross-referencing code, conversation history, and expert knowledge.

JavaScript Maintenance & Refactoring Standards

Enforces systematic standards for detecting technical debt, prioritizing refactoring tasks, and verifying JavaScript code quality.

Frontend Dashboard Components

Develops and maintains high-performance React 19 dashboard components using Next.js 15, shadcn/ui, and Recharts for sentiment monitoring interfaces.

CUI HTTP Testing Standards

Streamlines HTTP client testing in Java by enforcing CUI standards for MockWebServer and JUnit 5 integrations.

Guardian Quality Monitor

Monitors session health and automates code quality gates by spawning focused review agents when degradation or complexity thresholds are met.

TRL Training on Hugging Face

Trains and fine-tunes language models using Transformer Reinforcement Learning methods on fully managed Hugging Face cloud GPU infrastructure.

Reference and Fixture Manager

Maintains consistency and prevents duplication in AI coding tasks by managing and loading existing reference materials and test fixtures.

No-Reference Context Guard

Enforces strict context isolation by preventing AI reliance on external reference directories during code generation.

Meeting Minutes Synthesizer

Synthesizes meeting transcripts into detailed Confluence pages by capturing technical reasoning, trade-offs, and substantive decision-making context.