Desenvolvimento

Craft CMS 5 plugin and module development — extending Craft with PHP. Covers the full extend surface: elements, element queries, services, models, records, project config, controllers, CP templates, migrations, queue jobs, console commands, field types, native fields, events, behaviors, Twig extensions, utilities, widgets, filesystems, permissions, debugging, testing, GraphQL, and Craft configurat

Explore different architectural approaches in parallel using hosted LLM for code generation. No restrictions on approach - fastest path to comparing real implementations. Part of speed-run pipeline.

Coleta o resumo diário de tendências de IA e Tecnologia de shir-man.com e gera um belo relatório HTML com abas.

Previews the MorningStar task plan for a PRD without writing any code or making commits. Fetches the Notion PRD, analyzes the current codebase, and prints the proposed task list with files, acceptance criteria, and an estimated cost. Use when the user says any of "preview", "dry run", "what would this build", "show me the plan", "estimate cost", "morningstar dry-run", or wants to validate a PRD be

Docs-aware grilling: interrogates the user's plan one question at a time AND maintains the domain glossary (CONTEXT.md) and architecture decision records (docs/adr/) inline as decisions crystallise. Used to stress-test a plan against the project's language and documented decisions; for a plain interview that writes no files, use grill. Activates when the user writes "grill-docs", "grill with docs"

Token de sessão da awp-wallet; apenas a awp-wallet legada (<v0.17) precisa disso.

Use when the user describes work to be tracked and the work kind isn't obvious. Infers whether it's a task, story, bug, spec, epic, or incident from the language used, then routes to the right entry point (`/ticket`, `/spec`, or direct boardmaster call) with the correct `kind` pre-set.

Use when reviewing source files, diffs, commits, pull requests, or bugfixes for correctness issues, regressions, security problems, performance risks, edge cases, or missing tests across any language or framework.

Execute uma pergunta, decisão ou ideia através de um conselho agêntico de 5 assessores. Cinco subagentes debatem de perspectivas independentes em paralelo, com um Presidente emitindo o veredito após rodadas opcionais.

Esta skill aborda padrões de origem-como-impacto na derivação de agentes, explorando como eventos de separação criam agentes que carregam material inédito. Ela também examina como uma origem violenta pode gerar um serviço confiável, baseando-se na dinâmica de formação Theia-Lua.

Orients Claude Code at the start of a new session by reading CLAUDE.md, the latest DECISIONS.md entry, and CONTEXT.md, then giving a clear briefing: what the project is, what was done last session, current state, and exactly what to work on next. Use when someone says "catch me up", "where did we leave off", "what were we building", "new session", "what's the status", "I just cloned this repo", or

Use when the user announces non-trivial work (refactor, feature, bugfix, improvement) and no ticket has been created for it yet. Checks if a ticket already exists; if not, suggests creating one (or a parallel batch) before implementation starts. Tracks work that would otherwise vanish.