SKILL DIRECTORY
설치 수와 최근 흐름을 기준으로 필요한 Agent Skill을 찾아보세요.
skills.sh API 기준방금 확인
Distributed training orchestration across clusters. Scales PyTorch/TensorFlow/HuggingFace from laptop to 1000s of nodes. Built-in hyperparameter tuning with Ray Tune, fault tolerance, elastic scaling. Use when training massive models across multiple machines or running distributed hyperparameter sweeps.
Use when generated assets must keep a consistent style, character, or product look and references stop scaling, or when a user asks to train a custom model through the Scenario MCP, fine-tune a LoRA, clone a voice, curate a training dataset, choose a base model, set epochs and sample prompts, estimate training cost, diagnose a trained model that lost its identity or has one epoch, or generate with a trained model. Keywords: custom model, LoRA, dataset curation, epoch previews.
Analytics and live ops - AnalyticsService economy/funnel/onboarding/progression/custom events, player segments, experience configs via ConfigService, A/B experiments, badges, experience notifications, timed events, daily rewards, season passes. Use when instrumenting a game, tuning economy or onboarding, running A/B tests or feature flags, or planning live events and re-engagement.
Use when color grading, color correcting, or relighting an existing video or still with deterministic code instead of a generative model: cinematic looks (teal and orange, film stock, bleach bypass, noir, duotone, faded 70s), golden hour, blue hour or day-for-night, log footage (S-Log3) to Rec.709, matching a reference image's colors, black and white with one kept or swapped color, .cube LUT export, a look comparison grid, or a grading showreel.
Guides Qdrant deployment selection. Use when someone asks 'how to deploy Qdrant', 'Docker vs Cloud', 'local mode', 'embedded Qdrant', 'Qdrant EDGE', 'which deployment option', 'self-hosted vs cloud', or 'need lowest latency deployment'. Also use when choosing between deployment types for a new project.
Review a Kibana PR for OpenAPI Spec (OAS) and API docs compliance against the Elastic API docs checklist, core guidelines, and Kibana quickstart. Checks availability, summaries, descriptions, parameters, examples, tags, links, defaults, enums, deprecation, and generated YAML correctness. Use when reviewing a Kibana PR that touches API routes or OAS output files, or when asked to check OAS, availability, x-state, or API docs compliance on a PR.
Use when a named character or prop must become a reusable identity in Scenario: creating a character or prop from a brief or interview, building its approved image collection (character: <name> or props: <name>), assembling a character sheet or turnaround, or generating new images of an established one. Triggers include create a new character, design a prop, character sheet, turnaround, more images of this character. Keywords: character design, identity, library, collection, grid.
Build a MotherDuck dashboard as a Dive, choosing the analytical story, metrics, and section queries.
Reply to unresolved PR review comments that clearly relate to the current conversation. Use when the user asks to respond to PR review feedback without resolving threads or editing code.
Use when a mobile game's gameplay recording, screenshots or existing art and sound should become a playable ad: an HTML5 interactive ad or interactive end card for Meta, Google Ads App campaigns, AppLovin, Unity Ads, Mintegral, Moloco, Liftoff, TikTok Pangle or any MRAID network, with a hint hand, a CTA and a mute button, packaged per network and QA-tested in headless Chromium. Keywords: playable ad, HTML5 playable, MRAID, FbPlayableAd, ExitApi, end card, CTA, hypercasual ad.
Look up or repair DuckDB SQL syntax and verify MotherDuck-specific command and feature support.
>-
Build a scalable outbound B2B sales machine with specialized roles (SDR, AE, CSM). Use when the user mentions "outbound sales", "Cold Calling 2.0", "cold email sequences", "sales pipeline", "SDR process", "sales development", "build an outbound sales team", or "fill my pipeline". Also trigger when setting up a B2B SaaS sales team from scratch or building a lead-qualification framework to improve close rates. Covers the three lead types (seeds/nets/spears), role specialization, the referral-email method, ANUM qualification, and pipeline math. For offer design, see hundred-million-offers. For persuasion science, see influence-psychology.
Work out where a teammate, the champion, or a team is on the steps of AI adoption, and what to try next. Use when a champion wants to help someone move up a step.
>-
Fixes content and semantic HTML issues. Adds server-side rendering, a correct heading hierarchy, semantic landmark elements, ARIA labels, image alt text, the lang attribute, descriptive link texts, form labels, autocomplete tokens, keyboard access, document titles, button names, correct aria-hidden use and widget state so AI agents can read and operate a page through the HTML and the accessibility tree. Use when asked to "fix semantic HTML", "add SSR", "fix heading hierarchy", "add alt text", "improve semantics score", "fix accessibility for AI", "fix form labels", "add autocomplete", "fix keyboard navigation", "fix button names", "fix aria-hidden", "add aria-expanded", or any semantic HTML task.
Plans, executes, and troubleshoots migration from the Built-in Render Pipeline to URP. Use when asked to convert a project, scene, material, or shader to URP, or to fix visuals broken by the change.
Diagnoses Qdrant search quality issues. Use when someone reports 'results are bad', 'wrong results', 'not relevant results', 'missing matches', 'recall is low', 'approximate search worse than exact', 'which embedding model', 'should I fine-tune my embedding model', 'quality dropped after quantization', 'how to measure retrieval quality', 'build a golden set', 'ground truth dataset', 'how to score recall@k', or 'is my improvement real / statistically significant'. Also use when search quality degrades without obvious changes.
Build ingestion-to-serving pipelines on MotherDuck, including stage boundaries, transformations, and publication. Use for any recurring ingestion from an API, SaaS app, or database into MotherDuck.