深度學習

MCPサーバー設計パターンとアーキテクチャベストプラクティス。ツール組織化、エラーハンドリング、状態管理、サーバーライフサイクル管理の実証済みパターンを提供。Anchors:• Clean Architecture / 適用: サーバー構造と依存関係管理 / 目的: テスト可能で保守性の高いコード• Domain-Driven Design / 適用: ツールドメインモデリングと境界付きコンテキスト / 目的: 機能別ツール組織化• Pragmatic Programmer / 適用: エラーハンドリングと回復性パターン / 目的: 堅牢なサーバー構築Trigger:Use when designing MCP server architecture, organizing tool definitions, implementing error handling patterns, managing server state, structuring MCP server projects, or refactoring existing MCP servers.

Generate comprehensive multi-tenant security and operational reports from LimaCharlie. Provides billing summaries, usage roll-ups, detection trends, sensor health monitoring, and configuration audits across multiple organizations. Supports both per-tenant detailed breakdowns and cross-tenant aggregated roll-ups. Built with strict data accuracy guardrails to prevent fabricated metrics. Supports partial report generation when some organizations fail, with transparent error documentation. Time windows always displayed, detection limits clearly flagged, zero cost calculations.

This skill allows Claude to construct and configure neural network architectures using the neural-network-builder plugin. It should be used when the user requests the creation of a new neural network, modification of an existing one, or assistance with defining the layers, parameters, and training process. The skill is triggered by requests involving terms like "build a neural network," "define network architecture," "configure layers," or specific mentions of neural network types (e.g., "CNN," "RNN," "transformer").

This skill optimizes deep learning models using various techniques. It is triggered when the user requests improvements to model performance, such as increasing accuracy, reducing training time, or minimizing resource consumption. The skill leverages advanced optimization algorithms like Adam, SGD, and learning rate scheduling. It analyzes the existing model architecture, training data, and performance metrics to identify areas for enhancement. The skill then automatically applies appropriate optimization strategies and generates optimized code. Use this skill when the user mentions "optimize deep learning model", "improve model accuracy", "reduce training time", or "optimize learning rate".

Perform comprehensive project housekeeping - update roadmap, reconcile issues with implementation reality, organize completed work, and identify drift. This is a workflow skill that coordinates multiple parallel subagents for efficiency. Use when user says something like "run housekeeping", "do your housekeeping" or "clean up project state".

MIGRATE CLAUDE.md into modular `.claude/rules/` directory structure following Claude Code's rules system. Converts monolithic CLAUDE.md into organized, path-specific rule files with glob patterns. Use when migrating to rules system, modularizing project instructions, splitting CLAUDE.md, organizing memory files. Triggers on "migrate claudemd to rules", "convert claude.md to rules", "modularize claude.md", "split claude.md into rules", "migrate to rules system".

Analyze Firefox bookmarks and tags directly from the places.sqlite database. Extract recent bookmarks, analyze tag popularity, find bookmark statistics, and explore bookmark metadata. Use when working with Firefox bookmarks, analyzing bookmark usage, organizing tags, or extracting bookmark data from Firefox profiles.

Enterprise-ready Python development incorporating Kaizen (continuous improvement) and Monozukuri (meticulous craftsmanship) principles. Use this skill when building Python applications, APIs, CLI tools, data pipelines, automation scripts, or when the user requests clean, efficient, fast, simple, elegant, enterprise-grade, bulletproof, or production-ready Python code. This skill enforces modern Python 3.12+ best practices, type safety, testing patterns, security, and performance optimization.

Maintain focus points while exploring tangents when investigating requires depth. Use when: (1) explicit request to track original objectives while exploring related tangents or multiple investigation paths, (2) beginning investigative tasks where exploration leads through multiple related topics, (3) debugging interconnected problems requiring sequential exploration of branching causes, (4) analyzing unfamiliar systems where discovery reveals new paths that branch from initial focus.

Enforces project file organization standards from CLAUDE.md/PROJECT.md - auto-fix mode

Build node-based editors, interactive diagrams, and flow visualizations using Svelte Flow. Use when creating workflow editors, data flow diagrams, organizational charts, mindmaps, process visualizations, DAG editors, or any interactive node-graph UI. Supports custom nodes/edges, layouts (dagre, hierarchical), animations, and advanced features like proximity connect, floating edges, and contextual zoom.

Copilot agent that assists with project planning, scheduling, risk management, and progress tracking for software development projectsTrigger terms: project management, project plan, WBS, Gantt chart, risk management, sprint planning, milestone tracking, project timeline, resource allocation, stakeholder managementUse when: User requests involve project manager tasks.

Comprehensive project status evaluation using hive-mind coordination, GOAP planning, neural analysis, and AgentDB memory. Use when assessing architecture health, planning refactoring, or generating status reports.

Provides comprehensive project folder structure design guidelines and best practices. Defines standard directory organizations for diverse project types including monorepos, web frameworks, backend services, libraries, and extensions. Ensures scalable, maintainable architecture through consistent file organization patterns. Specializes in separation of concerns, modular architecture, and tooling integration.Use when: designing new project structures, organizing monorepo workspaces with tools like Turborepo/Nx, structuring NestJS backend projects, organizing React/Next.js frontend applications, designing Go service architectures, creating NPM package structures, organizing VSCode extension projects, structuring Chrome extension codebases, planning directory hierarchies, migrating legacy project structures, or establishing code organization conventions for teams.

Use during Phase 4 of spec creation to break specification into organized task groups with dependencies - analyzes requirements to determine natural groupings, presents structure for validation, and creates detailed actionable tasks

[Aesthetics] Maps aesthetic briefs to cultural references: art movements, photography, film, architecture, music, fashion. Outputs reference lists with visual language analysis. Organizes findings as Image and Memo nodes in Brain canvas.

This skill should be used when the user asks to review, consolidate, organize, or clean up memories. Triggers on requests like "review what you remember," "consolidate my memories," "what patterns do you notice," "organize your memory," "perform memory maintenance," or "clean up old memories." Also use proactively during long sessions to maintain memory hygiene and prevent context overload.

Reforge neural dendrite patterns after catastrophic failure.

Train and deploy neural networks in distributed E2B sandboxes with Flow Nexus

Provides Vitest-based TypeScript testing expertise and best practices. Ensures proper test structure, mocking strategies, async handling patterns, and coverage optimization. Specializes in unit testing, integration testing, test organization with describe/it blocks, vi.mock patterns, snapshot testing, and testing library integration for UI components.Use when: writing TypeScript test files (.spec.ts, .test.ts), structuring tests with describe/it/test blocks, implementing mocks with vi.mock/vi.spyOn, testing async code with async/await or promises, creating test fixtures and helpers, implementing snapshot tests, testing React components with Testing Library, measuring and improving code coverage, or configuring Vitest for TypeScript projects.