Explorar skills

Use when implementation is complete and you need to merge, create a PR, or clean up — verifies tests pass, presents 4 structured integration options, executes chosen workflow, and cleans up worktrees. Pairs with code-forge:worktree.

Activate a Julia project/environment for the session

Publish content to all configured platforms. Reads the content manifest, validates files, and invokes each platform's publish.sh with credential isolation. Supports --dry-run. Handles partial failures gracefully. Works independently — you can write markdown by hand and just use this skill.

Activates when user asks to plot, chart, or visualize data in the Julia REPL.

Pesquise no Apple Mail em todas as contas sincronizadas por palavra-chave, assunto, remetente ou qualquer combinação. Use esta skill para encontrar e-mails ou mensagens, com suporte a prefixos opcionais como remetente:, assunto: e filtros de tempo.

Gerar relatórios HTML autocontidos em vez de longas saídas de terminal. Para conclusões de tarefas, pesquisa, auditorias, comparações e bloqueadores.

Um especialista e estilista de moda pessoal para qualquer gênero, esta skill oferece recomendações de looks, ajuda a rastrear itens, constrói perfis de estilo, dá opiniões sobre peças e gera looks para ocasiões específicas. Também permite importar compras da Zalando ou Amazon e fornece sugestões de compras com links.

Update an existing Ansible role. Triggered by /update-role. Reads specific role files, applies the requested change following all global rules (FQCN, tags, no_log), shows a unified diff per file, waits for explicit user confirmation, then writes. Never overwrites silently.

Use when the user wants to produce all deliverables for an SAP Activate phase in one orchestrated run. This is the factory — it chains skills, dispatches agents in parallel, collects artifacts, and produces a phase-complete deliverable package with quality scoring.

Run skill benchmarks with discriminating-only assertions against evals.json for any model and any AI agent. Use when benchmarking a skill against a model not yet tested, running with_skill/without_skill eval pairs, producing benchmark-<model>.json, re-grading an existing run, adding Phase 2 model comparison results, reviewing results in the eval viewer, updating README benchmark tables, or cleanin

Use to draft a weekly newsletter. Produces a hook, deep dive, three quick reads, and a CTA, all humanized via Walter. Outputs word count, detection score, suggested subject line, and preview text.

When hardening a web application against Cross-Site Scripting (XSS) and injection attacks.