LLM & Agents

DO NOT invoke unless explicitly instructed. Core guidelines for orchestrating tasks with subagents.

Interpret Claude Code changelogs and generate user-friendly usage guides

Create, list, and remove git worktrees for parallel agent work. Use when you need isolated workdirs, new worktree branches, or safe cleanup workflows.

Generate multiple Vibes app variations in parallel with business models and rankings. Use when exploring different interpretations of a broad objective or loose creative prompt.

Operate a Home Assistant instance via the official REST/WebSocket APIs and backups, with safe plan/apply workflows for bulk, reviewable changes.

Multi-agent thread orchestration and coordination

Compensates for Claude's January 2025 knowledge cutoff by actively searching Apple's official iOS 26 documentation before making claims about iOS 26 APIs, frameworks, or features. This skill should be used when about to make statements about iOS 26 features, suggesting iOS 26-specific APIs, implementing iOS 26 functionality, or answering questions about iOS 26 capabilities. The skill provides multi-tier search workflows using Apple RAG, Sosumi, and web search tools with proper verification tagging.

Creates and maintains Agent Skills with effective triggers and progressive disclosure. Use when user requests to create a skill, generate a SKILL.md, build custom capabilities, or mentions "create skill", "new skill", or "skill configuration".

This skill should be used when designing agent orchestration, implementing firewall architecture, creating phase-based workflows, adding severity batching, structuring agent hierarchies, or implementing SOTA multi-agent coordination patterns. Provides production-ready patterns for scalable agent systems.

Guide for creating effective skills. This skill should be used when users want to create a new skill (or update an existing skill) that extends Claude's capabilities with specialized knowledge, workflows, or tool integrations.

Use when setting up shell environment for Claude Code - installs zsh, Oh My Zsh, Powerlevel10k theme, MesloLGS NF fonts, Claude CLI completions, and productivity aliases

Use when completing work on a task or feature. Prompts for code review before finishing. ALWAYS invoke this skill before saying work is done.

Generate code in SLOP (Symbolic LLM-Optimized Programming), a language designed forminimal human involvement in coding. Use when: (1) User asks for SLOP code, (2) Needstrong contracts and range types, (3) Creating code with typed holes for incrementalgeneration, (4) Generating efficient C code. SLOP uses S-expression syntax and transpiles to C.

Markdown linting and automated fixing using markdownlint-cli2. Use when Claude needs to: (1) Check markdown files for style issues, (2) Fix markdown formatting problems, (3) Ensure markdown follows best practices, (4) Validate markdown documents, or (5) Apply consistent markdown styling

Design and implement UI/UX changes using existing components and Tailwind tokens; ensure accessibility. EXCLUSIVE to ui-ux-designer agent.

Meta-skill for creating, refining, and managing Agent Skills. Use when needing to create new skills, improve existing skills, analyze skill performance, or teach the agent new capabilities. Enables self-improvement and knowledge capture.

Use when coordinating code reviews, agent reviews, and approval workflows across Jira and Harness

Coordinates other skills and agents. ALWAYS use this skill on startup.

Run Codex plan validation or implementation review as a quality gate for non-trivial tasks. Use after planning or implementation.

Gerador e executor de quiz sobre conhecimento RAG. Comandos: '/gerar-quiz-rag gerar' para gerar perguntas do banco RAG e salvar em JSON, '/gerar-quiz-rag iniciar' para executar quiz de 10 perguntas, '/gerar-quiz-rag status' para ver progresso. Claude consulta o banco SQLite RAG e gera perguntas diretamente.