linhay/harmony-next.skills

Offline HarmonyOS NEXT developer skill library for AI coding assistants. Covers API 12–26 declarations, ArkTS/ArkUI, NDK, DevEco Studio, emulator automation, debugging, publishing, and reproducible local tooling.

What it solves

It provides AI programming assistants (such as Gemini CLI, Claude Code, and Codex) with a specialized, offline knowledge base for HarmonyOS NEXT development. This allows AI agents to accurately reference ArkTS, ArkUI, NDK, and DevEco Studio documentation without relying solely on their general training data, which may be outdated or incomplete.

How it works

The project is structured as a "skill pack" that can be installed into various AI agent environments. Once installed, the agent follows a specific retrieval path (SKILL.md $\rightarrow$ KITS.md/TASK_MAP.md $\rightarrow$ INDEX.md) to find local documentation, API references, and templates. It includes snapshots of API 26.0.0 declarations and automation scripts for emulator diagnostics and UI testing.

Who it’s for

Developers using AI coding assistants to build applications for HarmonyOS NEXT who need precise API references and automation workflows for the DevEco Studio environment.

Highlights

  • Comprehensive API Reference: Includes ArkTS, ArkUI, and system module references, specifically including API 26.0.0 snapshots.
  • Tooling Integration: Provides guides and scripts for DevEco Studio emulator automation, HVD, hdc, and packet capture diagnostics.
  • Ready-to-use Templates: Includes a minimal "Empty Ability" project for smoke testing and rapid prototyping.
  • Multi-Agent Support: Compatible with Gemini CLI, Claude Code, Codex, and DeepSeek Harness (DSH).

Related

  • Project
    SciML/Evolutionary.jl

    A Julia package providing a comprehensive suite of of evolutionary and genetic algorithms for complex optimization problems.

  • Project
    LeapLabTHU/limit-of-RLVR

    A research project and evaluation framework that analyzes whether Reinforcement Learning with Verifiable Rewards (RLVR) expands an LLM's reasoning capacity or merely optimizes its sampling efficiency.

  • Project
    wanghetommy/ichartjs

    An agent-first, renderer-independent visualization runtime that allows AI agents to inspect data, plan, validate, and render interactive charts and diagrams.

  • Project
    DMontgomery40/deepseek-mcp-server

    An MCP server that integrates DeepSeek's V4.1 Flash API into MCP-compatible clients, enabling text and visual chat and file management.