LLM 與 Agent

This skill provides comprehensive knowledge for working with the Anthropic Messages API (Claude API). It should be used when integrating Claude models into applications, implementing streaming responses, enabling prompt caching for cost savings, adding tool use (function calling), processing images with vision capabilities, or using extended thinking mode.Use when building chatbots, AI assistants, content generation tools, or any application requiring Claude's language understanding. Covers both server-side implementations (Node.js, Cloudflare Workers, Next.js) and direct API access.Keywords: claude api, anthropic api, messages api, @anthropic-ai/sdk, claude streaming, prompt caching, tool use, vision, extended thinking, claude 3.5 sonnet, claude 3.7 sonnet, claude sonnet 4, function calling, SSE, rate limits, 429 errors

SpecWorkflowMcpとAI連携による仕様駆動開発(SDD)完全自動化スキル。 PRD→SPEC→実装タスク→品質検証→Miyabi連携までを1コマンドで完結。 kiroやspec-flowの考え方を統合し、Claude Sonnet 4による高品質な仕様生成とタスク分解を実現。 あらゆるソフトウェア開発プロジェクトに適用可能な汎用SDDプラットフォーム。

Production tracing and metrics for multi-agent workflows. Track agent decisions, tool calls, and performance without monitoring conversation content.

Initialize .claude/ structure and surgically add PopKit section to CLAUDE.md without overwriting. Detects conflicts, creates config, prompts for Power Mode. Use for new projects only - use analyze-project for existing.

Spawn a Claude worker to handle a GitHub issue. Fetches issue details, creates a feature branch, and works autonomously. Use when you want to delegate an issue to a worker.

Best practices for writing effective Task tool prompts that produce clear, structured outputs from sub-agents. Covers context engineering, instruction clarity, output format specification, and parallel delegation patterns. MUST BE USED PROACTIVELY when spawning sub-agents to ensure high-quality results.

Autonomous iterative development loop system. Repeatedly executes prompts until completion conditions are met. Supports multiple completion modes, checkpointing, cost tracking, and safety guardrails.TRIGGERS - Keywords: loop, autonomous, iterate, iteration, iterative, repeat, until done, keep going, autonomous mode, /loop, continuous, persist, persistent, overnight, unattended, batch processing.TRIGGERS - Phrases: "keep working on this", "until it's done", "iterate until", "run autonomously", "work on this overnight", "continuous improvement", "keep refining", "don't stop until", "loop until complete".

Use when executing implementation plans with independent tasks in the current session - dispatches fresh subagent for each task with code review between tasks, enabling fast iteration with quality gates

GitHub issue-based swarm coordination for intelligent task decomposition and progress tracking. Use for transforming issues into multi-agent tasks, automated triage, task breakdown, and issue lifecycle management.

Enforce homelab-proxmox workflow patterns including single source of truth documentation, prompt logging, git commits, environment variables, and phase tracking. Use this skill when completing tasks, updating docs/ or CLAUDE.md files, or when workflow compliance validation is needed. Provides AI-powered fixes for violations.

Solidity smart contract development with Foundry. Covers writing, testing, security, deployment, and upgrades. Triggers on .sol files, contract, pragma solidity, forge.

Expert assistant for managing AI prompts, features, and configuration in the KR92 Bible Voice AI system. Use when creating AI prompts, configuring AI features, managing prompt versions, setting up AI bindings, or working with AI pricing and models.

プロジェクトの .claude/settings.json に新しいフックを追加する。「フック作成」「新しいフック」「フックを作って」「フック追加」「hook 作成」「フックを追加したい」「新規フック」などで起動。ツール実行前後やセッションイベントで実行されるフックを設定。

Generate complete project foundations with personalized claude.md, Docker setup, and project structure. Supports Guided Setup or Quick Start.

Angular 21+ state management with signals, NgRx SignalStore, resource APIs, AsyncState patterns, and localStorage abstraction for modern reactive applications

Create marketing content optimized for both human readers and LLM discovery (GEO/AEO). Use when the user needs to write or improve marketing materials including landing page copy, tweet threads, launch emails, blog posts, or feature announcements. Automatically analyzes the user's codebase to understand product features and value propositions. Applies casual, direct brand voice and Generative Engine Optimization principles to maximize visibility in AI search results.

Use when request is vague, complex, or needs design exploration BEFORE implementation. Ask questions one at a time. Do NOT use for clear, mechanical tasks. Output: design ready for OpenSpec proposal.

Interactive hook creator for Claude Code. Triggers when user mentions creating hooks, PreToolUse, PostToolUse, hook validation, hook configuration, settings.json hooks, or wants to automate tool execution workflows.

Run PHPUnit tests and fix failures using intelligent agent delegation. Use when user requests to:- Run tests (full suite, specific path, or single file)- Fix failing tests- Analyze test failures- Check test coverageDelegates to runner agent (haiku) for execution and fixer agent (sonnet) for fixes.Automatically cycles between run and fix until tests pass or human intervention needed.

VeloxTS framework assistant for building full-stack TypeScript APIs. Helps with procedures, generators (velox make), REST routes, authentication, validation, and common errors. Use when creating endpoints, adding features, debugging issues, or learning VeloxTS patterns.