AIAny
AI Agent2026
Icon for item

k-skill

A curated collection of Korea-focused AI agent skills that enable agents to perform local tasks — ticket booking, public-data lookups, e-commerce and messaging integrations — via prebuilt connectors and an optional hosted proxy for API keys and fallbacks.

Introduction

Most general-purpose agent toolkits lack reliable, production-ready integrations for country-specific services. k-skill fills that gap for Korea: a large catalog of prebuilt "skills" (HTTP wrappers, conversational flows and handoffs) that let an AI agent complete real-world tasks such as SRT/KTX booking, Seoul bike lookups, public-data queries, e-commerce searches and KakaoTalk archive search without engineering each integration from scratch.

What Sets It Apart
  • Korea-first coverage and flows — includes dozens of domain-specific skills (rail/bus booking, public-data endpoints, government forms, local marketplaces) so agents can call vetted endpoints and follow local flows instead of brittle screen-scraping or bespoke adapters. This reduces engineering time and failure surface for Korea-facing automations.
  • k-skill-proxy with hosted fallback and BYOK patterns — a proxy layer injects upstream API keys when available and offers hosted fallback routes for many public APIs, balancing ease-of-use and credential separation. Proxy-related code is AGPL-scoped while the main repo is MIT, so deployment choices matter.
  • Agent-agnostic install and plugin support — installable via a single npx command, with marketplace/plugin support for Claude Code and compatibility with multiple coding agents. That makes it straightforward to add skills to existing agent setups or CLI workflows.
  • Wide catalog and guardrails — the repo groups skills with clear notes on which require user login/credentials and which use operator-hosted keys, plus security/secret policies and per-skill documentation to surface trade-offs for automation vs. manual handoff.
Who It's For and Trade-offs

Great fit if you’re building Korea-targeted conversational automations or prototypes and want ready-made integrations to accelerate agent capabilities (developers, hobbyists, agent integrators). It’s especially useful when you need concrete flows for booking, public-data lookups, or local e-commerce and want a quick path to a working agent.

Look elsewhere if you need global, non-Korea coverage as a primary requirement, or if your organization mandates a single permissive license for all components (the proxy pieces are AGPL). Also note operational trade-offs: some skills require the user to supply personal login/keys, others rely on operator-hosted fallbacks; Node.js 18+ is required for local use. Expect to review the security/secret guidance before enabling credentialed paths.

Information

  • Websitegithub.com
  • AuthorsNomaDamas
  • Published date2026/03/24

More Items

Hugging Face
AI Model2026

A 4B causal language model post‑trained from Qwen3.5‑4B for agentic workloads — tool use, coding and instruction following — using a routing‑harness feedback loop aimed at iterative capability improvement; distributed as text-only safetensors with native 262,144-token context.

Hugging Face
AI Model2026

Multimodal agentic model for long-horizon computer and browser tasks, with visual self-correction and function-calling. The Pro variant is a 397B Mixture-of-Experts (≈17B active) model with a 262,144-token context window, Docker deployment recipes, and weights currently marked “coming soon.”

Hugging Face
AI Model2026

A multimodal, agentic LLM optimized for long‑horizon, visually grounded workflows — capable of operating browsers and terminals and autonomously executing and testing code. Open‑source weights are available and the family ships in mini, Pro and Max variants for different compute/quality tradeoffs.