Pesquisa e Web

Extended documentation skill for Claude Code installations. Complements built-in /team-onboarding with multi-mode research — internal setup analysis, external architecture docs, and web research from forums/GitHub/marketplaces.

Use quando o usuário deseja baixar ou inspecionar áudio/vídeo online diretamente com o CLI `yt-dlp` instalado. Ideal para vídeos únicos, playlists, extração de áudio, download de legendas, inspeção de formato, modelos de saída, fluxos de cookies do navegador e solução prática de problemas de comandos `yt-dlp`.

Manages grepai semantic search (setup, status, start, stop, reindex, optimize, upgrade).

Use esta skill para coletar material de pesquisa, como vídeos do YouTube e artigos da web, em um notebook do NotebookLM para análise. Ela executa consultas e salva os resultados em markdown, combinando a busca do YouTube com yt-dlp e a pesquisa CLI do NotebookLM.

Runs evidence-based multi-agent debates with mandatory Discovery phase (codebase + web research). Three modes: Challenge picks best variant from options, Strategy produces independent proposals via deep analysis, Critic finds all weaknesses. Triggers: debate, challenge this, compare options, strategy analysis, critique, find weaknesses, second opinion, pros and cons, evaluate alternatives.

Use when: 需要进行市场调研数据收集、竞品信息采集、行业舆情监控、合规政策检查 Do NOT use when: 已有完整数据无需补充、仅需内部数据分析无需外部信息

Use when the user wants to improve a specific article under a fixed viral-scoring rubric, iterate on one article to raise its score, or run single-article scoring plus rewrite loops. Best for content optimization experiments where the scoring prompt stays fixed and the article text is the thing being changed.

Use quando o usuário deseja iterar sobre um sistema de pontuação de artigos virais, calibrar ou melhorar um prompt de pontuação com base em amostras rotuladas, ou executar experimentos de pontuação em lote em um conjunto fixo de artigos. Ideal para pesquisa de pontuação baseada apenas em prompt, onde os scripts do avaliador permanecem fixos e apenas a rubrica/prompt de pontuação deve evoluir.

Creates SPEC.md task specification through codebase research and user interaction. First step in brewcode pipeline before plan and start. Triggers: create SPEC.md, brewcode spec, write task specification, generate spec from requirements, task SPEC from description, feature spec document.

Deep, iterative, multi-source research that explores a topic from multiple perspectives, detects contradictions, and produces a comprehensive vault note with fact-level citations and a confidence map. Use this skill whenever the user says /deep-research, "deep research X", "research X thoroughly", "investigate X deeply", "I need to really understand X", or wants research that goes beyond a quick w

Activate for: persona, buyer persona, ICP, ideal customer profile, target customer, who is my customer, define audience, target audience, customer profile, buyer profile, build persona, create persona, segment audience, market segment, who should we target, customer research, customer interview, voice of customer. NOT for: lead scoring (use lead-scoring), prospect research (use prospect-research),

Swiss jurisdiction oracle — resolves federal vs. cantonal law applicability, identifies the correct court system for all 26 cantons, and routes cross-cantonal conflicts. Trigger when: a query names a Swiss canton (ZH, GE, BE, TI, VD, etc.) or canton court; when the user asks which court or law applies in a Swiss location; when federal vs. cantonal competence is disputed or unclear (e.g., 'does Zur