Skills

A PlantUML diagramming skill from the 365-Skills collection that enables creating UML, C4, sequence, and class diagrams within AI coding agents.

LaunchDarkly Experiments skill for setting up A/B tests with metrics, treatments, and data collection, enabling measurement-driven feature rollouts through LaunchDarkly's experimentation platform.

An AWS agent skill that guides connecting Lambda functions to API Gateway, part of the Agent Toolkit for AWS covering serverless architecture patterns.

LaunchDarkly feature flags skill for configuring guarded rollouts with progressive traffic ramp-up, metric monitoring for regressions, and automatic rollback capabilities to safely release features to production.

Production-grade web scraping and automation skills for AI coding agents, providing access to the Apify platform with 25,000+ Actors for scraping social media, search engines, maps, and review sites.

Screenshot optimization skill from ASO Skills suite. Guides creation of effective app store screenshots that convert browsers to downloads.

Part of PM Skills Marketplace, a collection of 65 product management skills and 36 chained workflows covering discovery, strategy, execution, launch, and growth, built for Claude Code and Cowork with structured PM frameworks.

Skill for guiding AI-assisted software building using insights from 3 product leaders, covering iterative prompting, goal breakdown, and expectation setting for prototypes and MVPs

A Korean public parking lot search skill that finds nearby public parking lots based on the user's current location using official Data.go.kr open data APIs. It geocodes location via Kakao Map, returns results sorted by distance with fees, hours, and map links, and always asks for location before searching.

AI product-photography skill that turns a product image or description into ad-ready photos in 5 styles (Studio, Floating, Ingredient, In Use, Lifestyle) at 1:1 and 9:16, via banana-claude's nanobanana MCP. Loads optional `brand-profile.json` for color/voice injection, picks Product vs Editorial mode per style, and files outputs under `./product-photos/<style>/`.

Apple Ads (formerly Apple Search Ads / ASA) deep-analysis skill for mobile app advertisers — evaluates campaign structure (Brand/Category/Competitor/Search Match), CPT/TTR/CR bid health, Custom Product Pages (Creative Sets deprecated), MMP/AdAttributionKit attribution, Maximize Conversions, and TAP coverage (Today/Search/Product Pages) into a 0–100 ASA Health Score with action plan.

A thin wrapper skill that generates Korean law-compliant privacy policies, terms of service, cookie banners, and consent modals for Next.js App Router projects, based on the Personal Information Protection Act, E-Commerce Act, and related Korean regulations.

A spec-driven development skill for Claude Code that maintains a durable SPEC.md using caveman encoding (~75% fewer tokens than prose), with automatic backpropagation that turns test failures into persistent spec invariants the agent never forgets.

A TanStack Start skill for the full-stack React framework with SSR, streaming, and server functions. Part of the tanstack-skills collection covering the complete TanStack ecosystem.

Designs comprehensive user interfaces, wireframes, and design systems with a focus on accessibility, user research, and modern design tools.

A documentation review skill from the Compound Engineering plugin, applying the principle of making each unit of engineering work easier than the last through thorough planning and review workflows.

Part of the LaunchDarkly agent skills collection, this skill migrates applications with hardcoded LLM prompts to LaunchDarkly AI Configs through a five-stage workflow: extract, wrap, tools, tracking, and evals.

An interactive grilling session that stress-tests your plan against the existing domain model, sharpens fuzzy terminology, cross-references code for contradictions, and updates CONTEXT.md and ADR documentation inline as decisions crystallize.

Part of the LaunchDarkly agent skills collection, this skill helps select the right metric type for experiments and create metrics with proper tracking event instrumentation for LaunchDarkly's experimentation platform.

A skill that searches Korean public library collections via the Data4Library Open API, enabling book keyword search, ISBN-based detail lookup, library holdings queries by region, and availability checks at specific libraries.

Part of PM Skills Marketplace, a collection of 65 product management skills and 36 chained workflows covering discovery, strategy, execution, launch, and growth, built for Claude Code and Cowork with structured PM frameworks.

Expert guidance for GitHub CLI (gh) operations covering pull request management, issue tracking, repository search, label management, GitHub Actions workflow triggering, codespace management, and release automation from the command line with built-in safety rules.

App Store Connect CLI skill for managing build processing, finding latest builds, distribution flows via TestFlight and App Store, and cleaning up old builds

Skill for removing sloppy or low-quality code from Cursor plugins, a marketplace of official plugins for popular developer tools including frontend, design, data science, code review, and documentation workflows.

Full lifecycle for Claude skills — draft, run eval prompts against the skill, review qualitative and quantitative results with `eval-viewer`, iterate on description/content, then benchmark at larger scale.

Audits AI-generated "vibe-coded" applications for the security mistakes LLMs consistently introduce — hardcoded secrets, secrets exposed via `NEXT_PUBLIC_`/`VITE_`/`EXPO_PUBLIC_` prefixes, broken Supabase RLS / Firebase rules / Convex auth, weak JWT/middleware/Server Action protection, missing rate limits, client-side price manipulation, unverified Stripe webhooks, insecure mobile token storage, unsafe LLM API keys/output, and SQL/ORM input flaws. Loads relevant `references/*.md` files only when the codebase uses that technology and reports findings ordered Critical → High → Medium → Low with file:line and before/after fixes.

from vercel/vercel-plugin

Runs Nx tasks and commands across projects in a workspace, providing intelligent task execution and management for Nx monorepos.

Build durable background tasks with Trigger.dev that run reliably with automatic retries, queuing, and observability. Covers task creation, triggering from backend code, schema validation with Zod, scheduling cron jobs, and implementing concurrency control for AI agents and async workflows.

ArXiv paper reading skill by Andrej Karpathy (nanochat) that fetches TeX source from arxiv URLs, unpacks, and locates the main .tex file for reading papers in their original source format