メニュー
このリポジトリの skills
SkillsMP は Sheshiyer/skill-clusters から 722 件の skill を収集しています。skill を開くとソースと詳細を確認できます。
Sheshiyer/skill-clusters収集済み skill 722 件中 40 件を表示しています。
Build production-ready Discord bots with Discord.js (JavaScript) and Pycord (Python) — gateway intents, slash commands, interactive components, rate limiting, and sharding. USE WHEN building, structuring, or scaling a Discord bot and choosing intents,…
原文の言語: 英語
Build evaluation frameworks for non-deterministic agent systems: rubric design, eval methodologies, test-set design, context-engineering eval, and continuous evaluation. USE WHEN you need to stand up an eval framework to validate or compare agent behavior.
原文の言語: 英語
Handle file uploads and cloud storage safely: S3, Cloudflare R2, presigned URLs, multipart uploads, content-type validation, and image optimization without blocking. USE WHEN an app or agent must accept user file uploads or stream large files to object…
原文の言語: 英語
Firebase backend expert across Authentication, Firestore, Realtime Database, Cloud Functions, Cloud Storage, Hosting, and security-rule design for denormalized read-heavy data. USE WHEN building or securing an app on Firebase/Firestore.
原文の言語: 英語
Verify that fix commits properly resolve security/audit findings at the root cause without introducing new bugs, regressions, or vulnerabilities, and confirm test coverage for the fixed scenario. USE WHEN reviewing remediation commits that claim to close…
原文の言語: 英語
Handle errors as values in TypeScript using fp-ts Either and TaskEither instead of try/catch, with validation-error accumulation. USE WHEN you want type-safe explicit error handling in TS APIs or services.
原文の言語: 英語
A practical, jargon-free 80/20 guide to fp-ts functional programming for handling nullable values, errors, and async in TypeScript without category-theory overhead. USE WHEN writing or refactoring TypeScript with fp-ts (Option/Either/TaskEither).
原文の言語: 英語
Practical fp-ts patterns for React — Option/Either/TaskEither/RemoteData for state, forms, and data fetching on React 18/19 and Next.js 14/15. USE WHEN building React apps with functional, type-safe state and fetch handling.
原文の言語: 英語
Route a web-UI task to the right skill among 18 frontend specialists — framework choice (React/Next, Vue/Nuxt, Angular), build tooling (Vite, Turbopack, Bun), motion, accessibility, design direction, and component conversion. USE WHEN building, reviewing, or…
原文の言語: 英語
Production-ready serverless on GCP — Cloud Run services (containerized), Cloud Run Functions (event-driven), cold-start optimization, and Pub/Sub event architecture. USE WHEN deploying or tuning Cloud Run containers/functions or wiring Pub/Sub-driven…
原文の言語: 英語
Route a growth/marketing task to the right skill among 13 demand-generation specialists — content strategy, long-form writing, conversion copy and editing, cold and lifecycle email, paid ad creative, lead magnets and free tools, marketing psychology, the idea…
原文の言語: 英語
Scoped CLAUDE.md memory system (HAM) that cuts context-token spend with directory-level context files, a decisions/patterns memory layer, and a savings dashboard. USE WHEN a multi-directory project keeps re-reading the same files and you want scoped agent…
原文の言語: 英語
Expert patterns for HubSpot CRM integration with the Node.js and Python SDKs — OAuth authentication, CRM objects, associations, batch operations, webhooks, and custom objects. USE WHEN integrating an app with HubSpot CRM, syncing contacts/deals, wiring…
原文の言語: 英語
Use the Hugging Face Hub CLI (hf) for authentication, downloading, uploading, and managing models, datasets, and Spaces. USE WHEN you need to authenticate to or move models/datasets/Spaces on the Hugging Face Hub via CLI.
原文の言語: 英語
Run local evaluations of Hugging Face Hub models with inspect-ai or lighteval, including backend selection and GPU smoke tests. USE WHEN evaluating a Hugging Face model on local hardware outside the Jobs workflow.
原文の言語: 英語
Query Hugging Face datasets through the Dataset Viewer API for splits, rows, full-text search, filters, and parquet links (read-only). USE WHEN you need read-only exploration or extraction from a Hugging Face dataset.
原文の言語: 英語
Build or edit Gradio apps, layouts, components, and chat interfaces in Python for ML demos and UI prototypes. USE WHEN a user wants a Gradio demo or Python ML interface for a model or agent.
原文の言語: 英語
Run workloads on Hugging Face Jobs with managed CPUs, GPUs, TPUs, secrets, and Hub persistence, with no local setup. USE WHEN you need to run a training or batch workload on managed Hugging Face cloud compute.
原文の言語: 英語
Train or fine-tune language models with TRL on Hugging Face Jobs, covering SFT, DPO, GRPO, and GGUF export with results saved to the Hub. USE WHEN fine-tuning an LLM (SFT/DPO/GRPO) on managed Hugging Face GPUs.
原文の言語: 英語
Publish and manage research papers on the Hugging Face Hub: create paper pages, link models/datasets, claim authorship, and generate markdown research articles. USE WHEN publishing, indexing, or linking a research paper on the Hugging Face Hub.
原文の言語: 英語
Read and analyze Hugging Face paper pages or arXiv papers using markdown and the papers API metadata. USE WHEN you need to read, summarize, or pull metadata for an arXiv or HF paper.
原文の言語: 英語
Track ML experiments with Trackio via Python logging, alerts, and CLI metric retrieval, syncing to Hugging Face Spaces dashboards. USE WHEN you want lightweight experiment-tracking dashboards for an ML training run.
原文の言語: 英語
Train or fine-tune vision models on Hugging Face Jobs for object detection, image classification, and SAM/SAM2 segmentation on managed GPUs. USE WHEN fine-tuning a detection, classification, or SAM segmentation model on HF Jobs.
原文の言語: 英語
Inngest expert for serverless-first background jobs, event-driven workflows, and durable execution without managing queues or workers. USE WHEN building background jobs, durable multi-step workflows, fan-out, or scheduled functions on Inngest instead of…
原文の言語: 英語
Build, run, and debug the current iOS project on a booted simulator via XcodeBuildMCP — UI interaction, screenshots, view inspection, and runtime log capture. USE WHEN you need to run an iOS app on a simulator, reproduce a bug, tap/type through a flow, or…
原文の言語: 英語
Create and edit JSON Canvas files (.canvas) with nodes, edges, groups, and connections per the JSON Canvas 1.0 spec. USE WHEN working with .canvas files or building visual canvases, mind maps, knowledge maps, or flowcharts (e.g. in Obsidian).
原文の言語: 英語
Langfuse expert for open-source LLM observability: tracing, prompt management, evaluation, datasets, and integrations with LangChain, LlamaIndex, and OpenAI. USE WHEN you need to trace, monitor, or debug an LLM app in production with Langfuse.
原文の言語: 英語
Build, refactor, or review SwiftUI macOS menubar (LSUIElement) apps with a Tuist-first workflow, stable launch scripts, and strict model/client/store/view boundaries. USE WHEN working on a menu bar utility built with Tuist — manifests, run scripts, or keeping…
原文の言語: 英語
Scaffold, build, sign, notarize, and package SwiftPM macOS apps into .app bundles without an Xcode project, including Sparkle appcast and Gatekeeper validation. USE WHEN shipping a macOS app from a Swift package — bootstrapping the folder, codesigning,…
原文の言語: 英語
Design short-term, long-term, and graph-based memory architectures for agents that persist across sessions, maintain entity consistency, and reason over accumulated knowledge. USE WHEN choosing and designing the right memory architecture for a persistent…
原文の言語: 英語
Self-hosted semantic memory for agents via MCP: save worklogs, decisions, and notes to Postgres + pgvector and recall them across sessions by meaning with local embeddings. USE WHEN an agent needs a self-hosted, keyword-free semantic memory shared across…
原文の言語: 英語
Indie-hacker playbook to launch a small focused SaaS fast — idea validation, MVP scoping, pricing psychology, launch channels, and growth to sustainable revenue. USE WHEN going from idea to paying customers in weeks as a solo or small founder and you want a…
原文の言語: 英語
Route a native Apple-platform task to the right skill among the iOS specialists — Swift 6.2 concurrency, SwiftUI architecture, actor persistence, protocol-based DI/testing, on-device FoundationModels, Liquid Glass design, and icon assets. USE WHEN a user is…
原文の言語: 英語
Integrate Supabase Auth with the Next.js App Router — browser/server clients, middleware session refresh, and the auth callback route. USE WHEN wiring Supabase login, sessions, or protected routes into a Next.js app.
原文の言語: 英語
Build and sell Notion templates as a real digital-product business — template design that sells, pricing, Gumroad/Lemon Squeezy setup, marketplaces, marketing, and support systems. USE WHEN productizing Notion templates (or similar digital products) into…
原文の言語: 英語
Create and edit Obsidian Bases (.base files) with views, filters, formulas, and summaries. USE WHEN working with .base files or building database-like table/card/list views over a note collection, or when a user mentions Bases, filters, or formulas in…
原文の言語: 英語
Drive the Obsidian CLI to read, create, search, append, and manage vault notes, or develop/debug Obsidian plugins and themes from the shell. USE WHEN a user wants command-line interaction with a running Obsidian vault or scripted note management.
原文の言語: 英語
Create and edit Obsidian Flavored Markdown with wikilinks, embeds, callouts, properties, and other Obsidian-specific syntax. USE WHEN writing or editing notes meant to render in an Obsidian vault, or when a task involves wikilinks, callouts, frontmatter,…
原文の言語: 英語
Senior PM agent spanning strategy, discovery, prioritization, execution, and metrics — 30+ frameworks (RICE, JTBD, Kano, North Star), 12 templates (PRDs, roadmaps), and 32 SaaS metrics with formulas. USE WHEN you need structured product-management analysis:…
原文の言語: 英語
Caching strategies for LLM prompts including Anthropic prompt caching, response caching, and Cache-Augmented Generation (CAG) to cut latency and cost. USE WHEN you want to reduce LLM cost/latency by caching prompt prefixes or responses.
原文の言語: 英語