1,969 件 ・ 関連度順
概要と使いどころ
Comprima modelos de linguagem grandes usando destilação de conhecimento de modelos professor para aluno. Use ao implantar modelos menores com desempenho retido, transferir capacidades do GPT-4 para modelos de código aberto ou reduzir custos de inferência. Aborda escalamento de temperatura, alvos suaves, KLD reversa, destilação de logits e estratégias de treinamento MiniLLM.
日本語の概要は準備中です。原文の説明を表示しています。
Treinar modelos de Mixture of Experts (MoE) usando DeepSpeed ou HuggingFace. Use ao treinar modelos em larga escala com computação limitada (redução de 5× em custos vs modelos densos), implementar arquiteturas esparsas como Mixtral 8x7B ou DeepSeek-V3, ou escalar capacidade de modelo sem aumento proporcional de computação. Cobre arquiteturas MoE, mecanismos de roteamento, balanceamento de carga, paralelismo de especialistas e otimização de inferência.
日本語の概要は準備中です。原文の説明を表示しています。
Load this skill when building any web feature, reviewing architecture decisions, or evaluating JavaScript dependencies. Under no circumstances build features that break completely when JavaScript is unavailable or fails. Absolutely always start with semantic HTML, layer CSS enhancements, and add JavaScript as the final, optional layer. Prioritize resilience and universal access over cutting-edge features.
日本語の概要は準備中です。原文の説明を表示しています。
[Code Quality] Use when you need to prove fix correctness with adversarial code proof traces — a skeptic tries to DISPROVE the fix and trace all related paths first, with confidence scoring and stack-trace-style evidence chains, so confidence is earned by surviving attack, not by confirming it works.
日本語の概要は準備中です。原文の説明を表示しています。
[Dev Data] Use when you need to implement or enhance test data seeders that simulate QC happy-path scenarios via application-layer commands. Flag: --mode=review reviews a target seeder (or the current changes / current work-context result) against every universal seed-data rule AND the project-specific seeder conventions — read-only, evidence-backed PASS/FAIL.
日本語の概要は準備中です。原文の説明を表示しています。
[Decision Support] Use when pressure-testing irreversible, high-stakes decisions with adversarial AI advisors.
日本語の概要は準備中です。原文の説明を表示しています。
[Utilities] Use when initializing or re-evaluating portable project context: project-config, project-reference docs, CLAUDE.md, AGENTS.md, universal agent rules, hookless setup. Triggers on: project init, init project, missing project-config, missing project docs, missing CLAUDE.md, missing AGENTS.md.
日本語の概要は準備中です。原文の説明を表示しています。
[Code Quality] Use when reviewing architecture compliance for layers, messaging, service boundaries, CQRS, repos, entity events, and data/consistency/tenancy boundaries. Universal architecture laws, coupling taxonomy and the anti-pattern catalog live in `.claude/docs/architecture-knowledge.md` (project docs always outrank it).
日本語の概要は準備中です。原文の説明を表示しています。
Telegram CLI for AI agents. Use when the user needs to read or search Telegram, send or edit a message, download media, organise Saved Messages, monitor conversations, or automate a Telegram task. Triggers on “check my messages”, “send a message”, “search Telegram”, “read unread”, “listen to chat”, “download from Telegram”, тэг сохранёнок, чаты, каналы, and @peer names.
日本語の概要は準備中です。原文の説明を表示しています。
Virtual QA Lead that turns requirements into review-ready, TestRail-importable test cases in one conversation. Use it whenever the user has requirements, a spec, a user story, or acceptance criteria (PDF, DOCX, XLSX, TXT, MD, or pasted text) and wants test cases, a test plan, a checklist, test coverage, a regression suite, or a TestRail/Qase/Zephyr export, even without saying "test cases" ("write tests for this spec", "what should we check here?"). Especially useful when they attach their team's existing test cases to match. When the spec describes an API (endpoints, status codes, schemas, auth, OpenAPI/Swagger), Casely also builds a runnable Postman collection, so use it for "API tests", "Postman collection", or "tests I can run in Newman/CI". Triggers on the same request in any language.
日本語の概要は準備中です。原文の説明を表示しています。
Break a plan, spec, or the current conversation into a set of tracer-bullet tickets, each declaring its blocking edges, published to the configured tracker — edges as text in one file per ticket locally, or native blocking links on a real tracker.
日本語の概要は準備中です。原文の説明を表示しています。
Turn the current conversation into a PRD and publish it to the project issue tracker — no interview, just synthesis of what you've already discussed.
日本語の概要は準備中です。原文の説明を表示しています。
Turn the current conversation into a spec and publish it to the project issue tracker — no interview, just synthesis of what you've already discussed.
日本語の概要は準備中です。原文の説明を表示しています。
Compact the current conversation into a handoff document for another agent to pick up.
日本語の概要は準備中です。原文の説明を表示しています。
WeChat article creator with Fu Sheng's writing style - conversational, story-driven, opinionated. Use this skill whenever the user mentions writing a WeChat article, drafting public account content, crafting a viral headline, creating article covers, writing in Fu Sheng's style, or any request involving 公众号, 微信文章, 爆款标题, or 傅盛风格 — even if they don't explicitly say "skill".
日本語の概要は準備中です。原文の説明を表示しています。
Adversarially probe this monorepo for security flaws at the right depth. Steers the red-team subagent — picks quick/standard/deep from the request, spawns it (fanning out for a wide deep sweep), collates findings into a dated report, and files security/sec:* issues for confirmed high/critical findings. Use when asked to "red-team", "try to hack this", "find security holes", "pentest this package/app", or as the brain behind the CI + daily security runs.
日本語の概要は準備中です。原文の説明を表示しています。
Automatic quality check on significant outputs via isolated sub-agent. Spawns reflection-agent for objective evaluation without conversation context bias.
日本語の概要は準備中です。原文の説明を表示しています。
Validate all JSON-LD scripts in Astro pages — check parsability, @context, @type, and Layout.astro auto-inject compatibility.
日本語の概要は準備中です。原文の説明を表示しています。
运用《关键冲突》(Crucial Conversations 系列) 的 CPR 思维法与安全问责框架,指导用户当他人违反承诺时开展有效面对。适用于用户要谈下属拖延/同事甩锅/老板反悔承诺、反复遇到同一问题却谈不拢、准备高难度问责对话脚本、以及设计承诺跟进机制时使用。
日本語の概要は準備中です。原文の説明を表示しています。
Production error → triaged, root-caused, repaired, and certified (tsc + rehearsal + reproduce-then-gone) fix PR for a human to merge — then confirm the error stops recurring. Never auto-merges.
日本語の概要は準備中です。原文の説明を表示しています。
Production error → triaged, root-caused, repaired, and certified (tsc + rehearsal + reproduce-then-gone) fix PR for a human to merge — then confirm the error stops recurring. Never auto-merges.
日本語の概要は準備中です。原文の説明を表示しています。
Convert any phased plan file, or a conversation about what to build, into labelled GitHub issues under one tracking epic mapping each issue to its source task. No plan file required. Don't use for writing plans, resolving issues, or triage.
日本語の概要は準備中です。原文の説明を表示しています。
Azure AI VoiceLive SDK for Java. Real-time bidirectional voice conversations with AI assistants using WebSocket. Triggers: "VoiceLiveClient java", "voice assistant java", "real-time voice java", "audio streaming java", "voice activity detection java".
日本語の概要は準備中です。原文の説明を表示しています。
Azure AI Voice Live SDK for .NET. Build real-time voice AI applications with bidirectional WebSocket communication. Use for voice assistants, conversational AI, real-time speech-to-speech, and voice-enabled chatbots. Triggers: "voice live", "real-time voice", "VoiceLiveClient", "VoiceLiveSession"...
日本語の概要は準備中です。原文の説明を表示しています。