Dena - Coleção de Skills Reutilizáveis
arpit786mtr-del/Dena
Skills reutilizáveis que automatizam workflows repetidos no Claude
Ver no GitHubRepositório de skills reutilizáveis para Claude que encapsulam workflows repetidos. Inclui análises de gaps de mercado que mineram dados de X, Reddit e sites de reviews para identificar oportunidades de produtos e gerar relatórios automaticamente. Cada skill segue uma estrutura padrão com documentação YAML, referências e exemplos.
Instalação
npx convez-ai arpit786mtr-del/Dena1 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
Casos de uso
- Pesquisa de gaps de mercado e oportunidades de produtos
- Análise de reclamações e requests de usuários em redes sociais
- Geração automatizada de conteúdo (carrossel Instagram, posts LinkedIn)
- Automação de workflows que rodam frequentemente
Veredito

Projeto bem estruturado que oferece um padrão prático para reutilizar workflows no Claude. O exemplo de market-gap-research é sólido, mas o repositório é ainda incipiente (uma skill ativa) — promissor para quem busca padronizar automações recorrentes.
README
A tradução para o português deste README ainda está na fila.
Dena
A growing collection of reusable Anthropic Skills I use across Claude Code, Claude Desktop, and Claude.ai. Each skill captures a workflow I run repeatedly so I don't reinvent it every time. Maintained by @Arpit (founder, Accelerate AI).
Skills in this repo
Skill market-gap-research what it does Mines X, Reddit, and review sites via Apify to surface product gaps and unmet feature requests, then ships an Instagram carousel / LinkedIn post / research report. when to use it Competitor research, finding SaaS wedges, complaint mining, "what do users wish X did?" More skills coming. Each lives in its own top-level folder.
Repo conventions
Every skill follows the same shape so they stay easy to read, install, and remix:
/ ├── SKILL.md # Required. YAML frontmatter (name, description) + body. ├── references/ # Optional. Cheatsheets, templates, deep-dives the SKILL.md links to. └── examples/ # Optional. Real outputs from running the skill — useful as test fixtures. Rules of the road:
• SKILL.md is the contract. Frontmatter has name and description. The description is "Use when..." — triggering conditions only, no workflow summary (Claude reads it to decide whether to load the skill).
• references/ is for reusable artifacts. API cheatsheets, HTML templates, prompt libraries — anything SKILL.md says "see references/foo.md for details." • examples/ is for proof. A real run of the skill, committed verbatim. Doubles as a test fixture and as a portfolio piece. • Skill names are kebab-case verbs or noun phrases (market-gap-research, not MarketGapResearch or gap_research). • One skill per folder, flat at the repo root. No nesting.
Installing a skill
Claude code
Clone once
git clone https://github.com/parikshit10/skills.git ~/parikshit-skills
Symlink the skills you want into your Claude Code skills dir
ln -s ~/parikshit-skills/market-gap-research ~/.claude/skills/market-gap-research Restart Claude Code. The skill will show up in your skills list and auto-trigger on matching prompts.
Claude Desktop / Claude.ai
Skills folders here can be uploaded as a project. Drag the skill folder into a Claude Project and the agent will load SKILL.md along with its references.
Contributing / using these yourself
These are personal skills tuned for my workflow, but most are general enough to fork. PRs welcome if you spot bugs in the workflow logic or want to add platforms / actor alternatives.
If you fork, swap out the voice/aesthetic notes in each skill (mine are tuned for @Arpit and @Arpitkumar).