Docs & Writing

product-on-purpose Bundle

Runs one of the three lightweight prioritization presets (Eisenhower urgent-important 2x2, MoSCoW Must/Should/Could/Wont bucketing, or Pareto vital-few cut) caveat-first. It leads with the fact that this is a bundle of three different operations stapled by the word "prioritization", not one move, and each is weakly evidenced, then produces the one preset the user named with the discipline a canned template lacks. Use only when an Eisenhower matrix, MoSCoW list, or Pareto chart is asked for by name; for the rigorous versions prefer think-theory-of-constraints, think-one-way-vs-two-way-door, or think-decision-option-review.

qq5855144 Bundle

Fetches the content of any URL and returns clean Markdown text. Suitable for scenarios where users share links and need to read, analyze, summarize, or translate web page content.

salesforcecommercecloud Skill

Updating user guides, CLI reference, and API documentation for the B2C CLI project. Use when adding or changing CLI command docs, writing JSDoc for TypeDoc generation, updating Vitepress sidebar config, or creating new guide pages.

willakins Bundle

Validate and address Codex review concerns on the current branch. Use when asked to triage Codex review findings, inspect review comments, decide which concerns are valid, fix only proven issues, and report each concern. Do not use for general code review, unrelated test failures, or creating new skills.

willakins Bundle

Create phase-based scope-of-work or remaining-work PDFs that read like polished implementation outlines rather than generic exports. Use when Codex needs to present validated issues, workstreams, priorities, recommended order, or definition-of-done in a styled PDF, especially for scope-of-work documents, project plans, implementation outlines, remaining-work summaries, or phase breakdowns.

willakins Bundle

Use this skill when asked to count or rank contributors in a GitHub repo by lines added/deleted, merged PRs, approved PRs, or recent/all-time activity using local git history plus GitHub PR metadata. Do not use for PR creation, CI repair, code review, recent PR listing, or generic git stats.

mingchen666 Bundle

Turn Bilibili videos that are already registered and parsed in MindSpace, or Bilibili opus/article posts, into evidence-linked Markdown learning notes. Use whenever the user asks to 提取、提炼、总结、整理 B站/Bilibili 视频、课程、动态、图文或专栏，保存学习笔记，结合评论区补充观点，或从已解析视频中查找带时间戳的内容和关键画面。Video tasks require an authorized mediaId from the current attachment context; this skill never installs or runs its own ASR models.

mingchen666 Bundle

OfficeCLI native PPTX deck design layer. Use when the user wants a polished, editable .pptx presentation for education, learning, office reporting, training, research briefs, proposals, courseware, review decks, or a custom creative deck. This skill does not replace officecli-skills; it chooses scene templates, visual systems, layouts, QA rules, and then uses officecli-skills with office_write to create the native PPTX.

mingchen666 Bundle

Use for writing, editing, polishing, or structuring lab reports and course experiment reports across disciplines. Handles cases with a provided template, raw data, images, notes, or only a brief requirement, and can produce DOCX or Markdown deliverables.

mingchen666 Bundle

This skill should be used when the user asks to "梳理知识", "整理笔记", "画知识框架", "总结要点", "提炼知识点", "一章总结", "知识清单", "帮我整理这章", "做个知识图谱", "总结重点", "画思维导图", "归纳知识点", "整本书的要点", "帮我提炼考点", or when the user provides textbooks, lecture notes, slides, or documents and wants structured knowledge outlines, key point extraction, chapter summaries, one-page review sheets, or knowledge maps.

l4ci Bundle

Use when the user wants to scan the macro-environment around a company, industry, market, or decision using PESTEL (Political, Economic, Social, Technological, Environmental, Legal). Fans out six parallel analyst subagents (one per factor) to identify and assess the external forces, then prioritizes them by impact and certainty and draws implications in a detailed markdown report. Triggers include "PESTEL", "PEST analysis", "PESTLE", "macro-environment", "external factors", "what trends affect this market".

l4ci Bundle

Use when the user wants to build or test a real strategy for an organization, unit, or initiative, or to check whether an existing "strategy" is actually one. Captures the challenge, fans out competing diagnostician subagents to name the crux from different angles in parallel, then synthesizes the sharpest diagnosis into a kernel (diagnosis, guiding policy, coherent action) and audits the result against Rumelt's four hallmarks of bad strategy in a detailed markdown report. Triggers include "strategy kernel", "good strategy bad strategy", "Rumelt", "what's our strategy", "diagnosis guiding policy", "is this a real strategy".

l4ci Bundle

Use when the user wants to reason a problem up from fundamentals instead of by analogy to how it is already done, challenge the assumptions baked into a conventional approach, or rebuild a solution from the ground up. Frames the received approach and the goal it serves, fans out parallel subagents that tear down inherited assumptions and establish what is actually bedrock, then reconstructs a solution from the surviving fundamentals and contrasts it with convention in a detailed markdown report. Triggers include "first principles", "first-principles thinking", "reason from scratch", "challenge assumptions", "why do we do it this way", "from the ground up".

l4ci Bundle

Use when the user is preparing for a negotiation (salary, deal, partnership, vendor, dispute settlement) and wants to go in ready rather than improvising. Fans out five parallel analyst subagents over the structure of principled, interest-based negotiation (Fisher and Ury, plus BATNA and ZOPA): your interests, their likely interests, your BATNA and walk-away, their likely BATNA and walk-away, and options for mutual gain with the objective criteria to invoke. Each marks every claim as fact or assumption with a confidence. A synthesis pass maps the zone of possible agreement, sets your target and walk-away, sequences the moves, and names the assumptions worth testing, into a detailed markdown report. Triggers include "prepare for a negotiation", "negotiation prep", "BATNA", "getting to yes", "salary negotiation", "how should I negotiate".

l4ci Bundle

Use when the user wants a Blue Ocean Strategy analysis of a business or product, or asks how to stop competing on the same factors as everyone else and create uncontested market space. Maps the industry's strategy canvas, then fans out four parallel subagents (eliminate, reduce, raise, create) to build a new value curve through value innovation, then synthesizes the blue-ocean move and its risks into a detailed markdown report. Triggers include "Blue Ocean", "value innovation", "ERRC", "four actions framework", "strategy canvas", "uncontested market", "stop competing on the same factors".

l4ci Bundle

Use when the user wants a Porter's Five Forces analysis of an industry or market, or asks to assess an industry's competitive structure, attractiveness, or profit potential. Fans out five parallel analyst subagents (one per force), each scoring its force against the standard determinants with real evidence, then synthesizes overall industry attractiveness and strategic implications into a detailed markdown report. Triggers include "Five Forces", "Porter analysis", "how attractive is this industry", "competitive structure of X".

l4ci Bundle

Use when the user wants a force field analysis of a proposed change, decision, or goal, or asks what is driving and what is holding a change back, or whether a change is feasible as things stand. Fans out finder subagents in parallel that hunt the driving forces and the restraining forces across people, organizational, technical, and external lenses and score each by strength, then synthesizes the scored force field, the net read, and the highest-leverage moves (favouring the weakening of restrainers) into a detailed markdown report. Triggers include "force field analysis", "Lewin", "driving and restraining forces", "what's holding this change back", "forces for and against", "change readiness".

l4ci Bundle

Use when the user wants to diagnose a team's psychological safety and prescribe how to improve it, or asks why people are not speaking up, why mistakes get hidden, or why learning has stalled. Fans out parallel assessor subagents across the dimensions of the climate (voice, response to mistakes, help-seeking, inclusion of difference, risk-taking and challenge), each scoring the climate against Amy Edmondson's behavioral markers with evidence and confidence, then places the team on the safety-by-accountability 2x2, names the weakest dimensions and the behaviors driving them, and prescribes leader and team interventions in a detailed markdown report. Triggers include "psychological safety", "people aren't speaking up", "mistakes get hidden", "Edmondson", "safe to take risks", "fearless team".

l4ci Bundle

Use when the user faces a decision with uncertain outcomes and wants the choice that maximizes expected value, or asks to map options and chance events into a decision tree and solve it. Frames the root decision, its later choices, and the chance events into one tree, fans out parallel subagents to estimate each branch's probabilities and payoffs, then folds back (expected value at chance nodes, best choice at decision nodes) to mark the optimal path, with break-even and value-of-information analysis, in a detailed markdown report. Triggers include "decision tree", "expected value", "decision analysis", "EV", "fold back", "decision under uncertainty".

l4ci Bundle

Use when the user wants to take a Big Five (OCEAN) personality test or asks to assess their personality across Openness, Conscientiousness, Extraversion, Agreeableness, and Neuroticism. Administers the public-domain 50-item IPIP questionnaire conversationally, scores it, and writes a detailed markdown report. Triggers include "OCEAN test", "Big Five", "personality assessment", "what's my personality type".

l4ci Bundle

Use when the user wants a HEXACO personality assessment, or a Big Five test that also measures Honesty-Humility. Administers the public-domain 240-item IPIP-HEXACO questionnaire conversationally across six factors (Honesty-Humility, Emotionality, eXtraversion, Agreeableness, Conscientiousness, Openness) and their 24 facets, scores it, and writes a detailed markdown report. Triggers include "HEXACO test", "HEXACO personality", "Honesty-Humility", "Big Five plus honesty", "six-factor personality assessment".

l4ci Bundle

Use when the user wants a Value Proposition Canvas for an offering and a customer segment, or asks to test whether a product fits what a customer actually needs. Fans out six parallel analyst subagents, three describing and ranking the Customer Profile (jobs, pains, gains) and three describing the Value Map (products and services, pain relievers, gain creators), each marking claims as fact or assumption, then ranks the customer side and assesses fit: which top jobs, pains, and gains the value map addresses, which it misses, and which relievers and creators chase things the customer does not care about. Produces a detailed markdown report. Triggers include "Value Proposition Canvas", "VPC", "value proposition design", "problem-solution fit", "does this fit the customer".

l4ci Bundle

Use when the user has a list of initiatives, ideas, features, or projects and wants to prioritize them by impact and feasibility (the impact-effort matrix), sorting into quick wins, major projects, fill-ins, and thankless tasks. Defines the axes, scores each option in parallel, places them in quadrants, and recommends a sequenced plan in a markdown report. Triggers include "impact vs feasibility", "impact-effort matrix", "prioritize these", "what should we do first", "quick wins", "prioritization matrix".

l4ci Bundle

Use when the user wants to understand why a person, role, or team has gone disengaged, demotivated, resistant, or burnt out, and what would actually re-engage them, including assessing their own motivation. Runs Deci and Ryan's Self-Determination Theory by fanning out three parallel analyst subagents, one per basic psychological need (Autonomy, Competence, Relatedness), each scoring both satisfaction and frustration for its need with evidence and how much it bears on the problem, then names the binding need, places motivation on the controlled-to-autonomous continuum, and prescribes targeted need-supportive moves in a detailed markdown report. Guards against the reflex of blaming generic "low motivation" when one specific need is being thwarted. Triggers include "self-determination theory", "SDT", "autonomy competence relatedness", "intrinsic motivation", "why is the team disengaged", "basic psychological needs".

phuc-nt Bundle

Analyze codebase and manage project documentation. Use for doc initialization, updates, summaries, codebase analysis.

phuc-nt Bundle

Search library/framework documentation via llms.txt (context7.com). Use for API docs, GitHub repository analysis, technical documentation lookup, latest library features.

robsonrung Bundle

Control cmux workspaces, panels, terminal input, notifications, and sidebar state through its CLI. Use when the user asks to inspect or manage cmux, create or target a workspace or panel, send text or keys, or report task status in cmux.

sandeco Bundle

Cria um deck Mira completo em uma chamada, com uma folha por slide executada em paralelo por um workflow do Claude Code. Use quando o usuário pedir /mira-fast, deck rápido, modo turbo, deck inteiro de uma vez ou deck em paralelo. Aceita os formatos mira, mira-studio, mira-studio-full e mira-vertical. Para um slide avulso, use mira-animator.

sandeco Bundle

Slide de QUIZ AO VIVO no Mira: a plateia escaneia um QR-code e responde uma múltipla escolha num Google Forms; o slide lê as respostas via Google Sheets gviz/JSONP. Diferente do mira-survey, tem resposta correta, revelação controlada pelo apresentador e percentuais que só aparecem após revelar. Use SEMPRE que o usuário disser /mira-quiz, quiz ao vivo, pergunta com resposta correta, revelar resposta, ranking do quiz, quiz com QR, quiz tipo Mentimeter, quiz tipo Slido, ou pedir um slide onde a plateia responde pelo celular e a resposta correta é revelada no palco. Para enquete sem resposta correta use /mira-survey; para QR simples use /mira-qrcode.

sandeco Skill

Insere num slide do Mira (novo ou existente) um QR code escaneável, grande e centralizado, gerado de um link ou texto. O QR é gerado LOCALMENTE (pacote npm qrcode) e embutido como SVG inline: funciona por file://, sem API nem CDN em apresentação. Card limpo no padrão mira-3d (só o título, no máximo 6 palavras, e o QR central, sem legenda do link); módulos escuros sobre cartão branco, laranja só na moldura e no título. Herda a Regra Zero do mira-animator, mas o QR fica ESTÁTICO e o loop interno vai na moldura. Use SEMPRE que o usuário disser /mira-qrcode, QR code, qrcode, código QR, gera um QR, coloca um QR no slide, QR do link, QR de inscrição, QR de CTA, código para escanear, ou pedir um slide com um link escaneável.

sandeco Skill

Gera um deck VERTICAL 9:16 (1080x1920) pronto para gravar no OBS Studio, onde cada slide declara um de três layouts: camera (webcam ao vivo em tela cheia), split (animação quadrada estilo mira-squared em cima + câmera embaixo) e full (animação vertical estilo mira-vertical). A câmera do apresentador é embutida ao vivo no slide via getUserMedia (módulo mira/mira-camera.js), sem chroma key: no OBS basta capturar a janela. Sem câmera, a área vira verde chroma #00FF00 como plano B. Use SEMPRE que o usuário disser /mira-studio, deck para OBS, deck com câmera, slide com câmera, meio a meio, câmera embutida, deck para gravar vídeo comigo falando, ou pedir slides que misturam câmera e animação num vídeo vertical.

sandeco Bundle

Cria um slide de ENQUETE AO VIVO: a plateia escaneia um QR-code, vota num Google Forms, e o resultado se atualiza em TEMPO REAL no slide (donut 3D girando ou barras). Recebe dois links — o de votação (vira o QR gerado local) e o da planilha de respostas (o slide lê a contagem via endpoint gviz/JSONP). Use SEMPRE que o usuário disser /mira-survey, enquete, enquete ao vivo, votação em tempo real, poll, resultado ao vivo no slide, QR de enquete, a plateia vota e o slide atualiza, tipo Mentimeter, tipo Slido, ou pedir um slide que mostra votação em tempo real. Para um QR sem votação use /mira-qrcode; para um gráfico estático use /mira-chart.

sandeco Skill

Reenquadra um deck do Mira na regra dos terços, sem mudar a proporção: joga título + animação nos 2/3 da esquerda de um grid 3x3 e deixa a coluna da direita em cinza #333, limpa, para você sobrepor texto, lower-third ou câmera na edição. A animação é reformulada por metáfora para preencher os 2/3; não toca no original (cria arquivo -thirds ao lado). Aplica sobre 16:9, 1:1 (mira-squared) ou 9:16 (mira-vertical), ou cria slides do zero; lado cinza à direita por padrão, com opção de inverter. Use SEMPRE que o usuário disser /mira-thirds, regra dos terços, rule of thirds, composição em terços, anima à esquerda, deixa a direita livre, coluna livre, dois terços à esquerda, grid 3x3, espaço para texto/câmera ao lado, abre espaço à direita, cria um slide em terços, ou pedir a animação encostada num lado e o outro terço livre em cinza.

sandeco Skill

Extrator de contexto do Mira: lê uma fonte vinculada no mira.config.json (pasta de projeto, PDF, LaTeX ou texto) e produz um briefing estruturado para o /mira-planner. Use SEMPRE que o usuário pedir para criar slides "sobre" um projeto ou documento, mencionar uma fonte vinculada pelo nome, ou disser "extrai o conteúdo de", "lê o projeto", "prepara o material do", "briefing do projeto".

sandeco Bundle

Transforma o TEXTO de uma história (normalmente infantil) num deck do Mira que é a animação completa dela: cenas contínuas com câmera, closes por dissolve, clima e paleta que mudam com a emoção, atores em SVG de fonte aberta, legendas de conto com narração em voz (edge-tts), música de fundo, capa com QR code e servidor na rede local para rodar no celular ou tablet em tela cheia. O modelo só escreve DADOS (mira/historia.js) num vocabulário fechado; o runtime versionado mira-history.js anima. Use SEMPRE que o usuário disser /mira-history, "anima essa história", "história animada", "conto infantil animado", "transforma essa história em animação", "historinha para o celular", "contar essa história com narração", ou colar um conto e pedir animação. NÃO é para explicar conceito (isso é /mira-animator e /mira-sequence-director) nem para deck de slides comum.

sandeco Bundle

Mural de POST-ITS AO VIVO no Mira, em dois slides: no primeiro a pergunta com um QR-code, no segundo cada resposta em TEXTO LIVRE da plateia vira um post-it que cola na parede, um a um. Diferente do mira-survey e do mira-quiz, aqui a resposta é aberta, não há alternativas nem resposta correta, e nada é agregado: cada pessoa aparece com a frase dela. Use SEMPRE que o usuário disser /mira-post-it, post-it, mural de respostas, parede de respostas, resposta aberta, resposta em texto livre, pergunta aberta para a plateia, as respostas vão aparecendo no slide, a plateia escreve e aparece na tela, mural ao vivo, ou pedir um slide onde a sala escreve uma frase curta pelo celular e ela surge na projeção. Para votação em alternativas use /mira-survey; para pergunta com resposta correta use /mira-quiz; para um QR sem leitura de respostas use /mira-qrcode.