npm package discovery and stats viewer.

Discover Tips

  • General search

    [free text search, go nuts!]

  • Package details

    pkg:[package-name]

  • User packages

    @[username]

Sponsor

Optimize Toolset

I’ve always been into building performant and accessible sites, but lately I’ve been taking it extremely seriously. So much so that I’ve been building a tool to help me optimize and monitor the sites that I build to make sure that I’m making an attempt to offer the best experience to those who visit them. If you’re into performant, accessible and SEO friendly sites, you might like it too! You can check it out at Optimize Toolset.

About

Hi, 👋, I’m Ryan Hefner  and I built this site for me, and you! The goal of this site was to provide an easy way for me to check the stats on my npm packages, both for prioritizing issues and updates, and to give me a little kick in the pants to keep up on stuff.

As I was building it, I realized that I was actually using the tool to build the tool, and figured I might as well put this out there and hopefully others will find it to be a fast and useful way to search and browse npm packages as I have.

If you’re interested in other things I’m working on, follow me on Twitter or check out the open source projects I’ve been publishing on GitHub.

I am also working on a Twitter bot for this site to tweet the most popular, newest, random packages from npm. Please follow that account now and it will start sending out packages soon–ish.

Open Software & Tools

This site wouldn’t be possible without the immense generosity and tireless efforts from the people who make contributions to the world and share their work via open source initiatives. Thank you 🙏

© 2026 – Pkg Stats / Ryan Hefner

@melody-core/skills-openai

v0.1.0

Published

Node.js 版 OpenSkills - Agent Skill 渐进式披露架构, 基于 openai-node 扩展实现skills management

Readme

melody-skills-openai-core

基于 openaiNode.js 版 Agent Skill 框架,实现渐进式披露(Progressive Disclosure)架构。

特性

  • 三层渐进式披露

    • Layer 1(元数据):始终加载,用于技能发现与匹配
    • Layer 2(指令):按需加载,技能被选中时加载
    • Layer 3(资源):按条件加载 References 与 Scripts
  • SKILL.md 格式:基于 Markdown + YAML frontmatter 的技能定义

  • 智能 Reference 加载:explicit / implicit / always 三种模式,可配合 LLM 选择

  • 自动发现:从 references/ 目录自动发现参考文档

  • 脚本执行:支持通过 [INVOKE:name] 触发脚本(Node 本地执行)

  • OpenAI 兼容:使用 OpenAI API(或兼容接口)

  • 自动技能调用:根据用户查询匹配并注入对应技能

安装

npm install @melody-core/skills-openai

快速开始

使用 SkillAgent(推荐)

const { createAgent } = require('@melody-core/skills-openai');

async function main() {
  const agent = await createAgent({
    skillPaths: ['./skills'],
    apiKey: process.env.OPENAI_API_KEY,
    model: 'gpt-4',
  });

  const response = await agent.chat('帮我总结会议');
  console.log(response.content);
  console.log('使用的技能:', response.skillUsed);
}

main();

使用 SkillManager(底层 API)

const path = require('path');
const { SkillManager } = require('@melody-core/skills-openai');

async function main() {
  const manager = new SkillManager({
    skillPaths: [path.resolve('./skills')],
  });

  await manager.discover();
  const matched = manager.match('summarize meeting');

  if (matched.length > 0) {
    await manager.loadInstruction(matched[0].metadata.name);
    console.log(matched[0].instruction?.content);
  }
}

main();

技能目录结构

my-skill/
├── SKILL.md           # 技能定义(必需)
├── references/        # 参考文档
│   └── handbook.md
└── scripts/          # 可执行脚本
    └── upload.js

SKILL.md 格式

---
name: meeting-summary
description: 用结构化格式总结会议
version: 1.0.0
triggers:
  - "summarize meeting"
  - "会议总结"

references:
  - path: references/finance-handbook.md
    condition: "当内容涉及财务或预算时"
    mode: explicit
  - path: references/safety-policy.md
    mode: always

scripts:
  - name: upload
    path: scripts/upload.js
    description: 将总结上传到云存储
---

# 会议总结技能

你是一名专业的会议助手...

Reference 加载模式

| 模式 | 行为 | |------------|------| | explicit | 带条件,由 LLM 判断是否满足条件后加载 | | implicit | 无条件,由 LLM 决定是否对当前查询有用(默认) | | always | 始终加载(如安全规范、规格说明) |

环境变量

export OPENAI_API_KEY=your-api-key
export OPENAI_BASE_URL=https://api.openai.com/v1  # 可选
export OPENAI_MODEL=gpt-4  # 可选

API 概览

  • createAgent(options):创建并初始化 SkillAgent
  • SkillAgent:自动根据查询选择技能并对话
    • initialize():发现技能
    • chat(content, options):发送消息并得到 AgentResponse
    • selectSkill(name) / deselectSkill():手动选择/取消技能
    • reset():清空对话上下文
  • SkillManager:技能发现、加载、匹配、执行脚本
    • discover():发现所有技能(仅 Layer 1)
    • match(query, limit):匹配技能
    • loadInstruction(name):加载指令(Layer 2)
    • loadReference(skillName, refPath):加载参考(Layer 3)
    • executeScript(skillName, scriptName, options):执行脚本
  • OpenAICompatClient:OpenAI 兼容的 LLM 客户端

与 Python 版 OpenSkills 的对应关系

| OpenSkills (Python) | melody-skills-openai-core (Node) | |-------------------------|-----------------------------------| | create_agent() | createAgent() | | SkillAgent | SkillAgent | | SkillManager | SkillManager | | OpenAICompatClient | OpenAICompatClient | | AgentResponse | AgentResponse | | SKILL.md / references/ | 相同目录与 frontmatter 约定 |

当前 Node 版未实现:沙箱执行(AIO Sandbox)、流式 chat_stream、Reference 摘要与召回等高级特性,可按需在后续迭代中补齐。

示例

仓库内提供示例技能与运行脚本:

# 先构建
npm run build
# 设置 API Key 后运行
export OPENAI_API_KEY=your-api-key
node examples/demo.js "帮我总结会议:今天讨论了 Q1 目标..."

示例技能目录:examples/skills/meeting-summary/(SKILL.md + references/ + scripts/)。

License

Apache-2.0