convez.ai
SkillAutomaçãoAvançado
sfrangulov

Coleção de Skills para Agentes IA

sfrangulov/skills

Skills modulares para agentes: consulting, OSS selection, demo videos, research

Ver no GitHub
InicialPluginScore · de 0 a 100

Repositório com skills instaláveis via skills.sh que estendem capacidades de agentes IA. Inclui frameworks de consulting, seleção de produtos open-source, geração de vídeos de demo automatizada e pipelines de pesquisa estruturada com verificação de fontes.

Instalação

$npx convez-ai sfrangulov/skills

1 comando no terminal: detecta o formato do repositório (marketplace ou plugin avulso) e instala via Claude Code. Requer Node 18+ e o Claude Code instalado.

Métricas

Estrelas
2
Forks
0
Acompanhando
0
Saúde
42%
Último commit
há 4 meses
Versão
—
Issues abertas
0
Contribuidores
1

Casos de uso

  • Estruturar e resolver problemas complexos de negócios com framework MECE
  • Automatizar pipeline de pesquisa com verificação de fontes e provenance
  • Gerar vídeos de demonstração de apps React/HTML de forma automatizada

Veredito

Coleção profissional de skills bem arquitetados para tarefas complexas (consulting, pesquisa, produção de vídeos). Exige domínio técnico e ambiente com ferramentas específicas (Node 20+, ffmpeg, APIs), não é plug-and-play.

README

A tradução para o português deste README ainda está na fila.

sfrangulov/skills

A collection of agent skills installable via skills.sh.

Skills

  • consulting-problem-solving — interactive, user-led consulting framework for solving complex business problems. 8 steps from problem definition through MECE structuring, prioritization, analysis, synthesis, and final deliverables (slides or vertical document).
  • consulting-problem-solving-ru — русскоязычная версия Consulting Problem Solving Framework. Тот же 8-шаговый процесс, переведённый на русский язык, с русскими триггер-фразами для активации.
  • oss-product-selection — structured workflow for picking the next open-source product to build when the goal is personal brand growth (not revenue). 7 phases with hard gates: constraints, asset inventory, broad ideation, competitive/platform/pain/hype scans, two-axis brand-vs-product scoring, naming, reality check and spec.
  • demo-video-pipeline — end-to-end automated pipeline for recording polished 30–90s demo videos of React/HTML apps. Universal zoom-hook with keyboard scenes → Playwright headless recording → Remotion overlay graphics → ElevenLabs voiceover with frame-perfect timings → final mp4 with audio. Requires Node 20+, pnpm, ffmpeg; ElevenLabs API key optional (web UI works too).
  • research-pipeline — disciplined multi-source research pipeline producing a verifiable, snapshot-backed document. Orchestrated subagents with structured specs, source-discipline, a fetch-escalation contract, a deterministic→judge verification funnel, content-addressed snapshots + provenance manifest, and a mandatory adversarial pass. Bundled Python tools, no API keys. Explicit-invocation only.

Install

Install a single skill:

npx skills add sfrangulov/skills --skill consulting-problem-solving

Install the Russian version:

npx skills add sfrangulov/skills --skill consulting-problem-solving-ru

Install all skills from this repo:

npx skills add sfrangulov/skills --skill '*'

Install several at once:

npx skills add sfrangulov/skills --skill consulting-problem-solving --skill another-skill

License

MIT — see LICENSE.

Ver o original no GitHub

Bom para

Relacionados