← 返回 Skills 市场
panb-kg

Dobby Harness Self-improving Coding Skills

作者 Panb-KG · GitHub ↗ · v1.0.0 · MIT-0
cross-platform ✓ 安全检测通过
60
总下载
0
收藏
0
当前安装
1
版本数
在 OpenClaw 中安装
/install dobby-harness
功能描述
Provides multi-agent orchestration with task decomposition, parallel execution, result aggregation, and self-improving workflows for complex coding tasks.
使用说明 (SKILL.md)

Dobby-harness Skill

多 Agent 编排 · 生产级工作流 · 自进化系统

📖 技能描述

Harness Engineering 提供完整的多 Agent 编排能力,包括任务分解、并行执行、结果聚合、自进化系统等功能。适用于需要协调多个子 Agent 完成复杂任务的场景。

🎯 适用场景

当以下情况时使用此 Skill:

  1. 复杂任务分解 - 需要将大任务分解为多个可并行执行的子任务
  2. 多 Agent 协作 - 需要协调多个专业 Agent 协同工作
  3. 生产工作流 - 需要代码审查、测试生成、文档自动化等标准化流程
  4. 状态持久化 - 需要防止上下文丢失、支持崩溃恢复
  5. 知识沉淀 - 需要从任务中学习并沉淀最佳实践

🚀 快速开始

基础用法

import { HarnessOrchestrator } from 'dobby-harness/harness/orchestrator.js';

// 创建编排器
const orchestrator = new HarnessOrchestrator({
  maxParallel: 5,
  timeoutSeconds: 300,
});

// 执行并行任务
const result = await orchestrator.execute({
  task: '分析项目代码质量',
  pattern: 'parallel',
  subTasks: [
    { task: '检查代码风格', agent: 'linter' },
    { task: '检查安全漏洞', agent: 'security' },
    { task: '检查性能问题', agent: 'performance' },
  ]
});

console.log(`完成:${result.completed}/${result.total}`);

使用工作流

import { CodeReviewWorkflow } from 'dobby-harness/workflows/code-review.js';

const workflow = new CodeReviewWorkflow();

const result = await workflow.execute({
  prNumber: 123,
  files: ['src/auth.js', 'src/user.js'],
  autoComment: true,
});

console.log(`审查评分:${result.report.score * 100}`);

📚 核心组件

1. Harness Orchestrator

多 Agent 编排核心,支持 5 种任务分解模式:

模式 说明 适用场景
parallel 完全并行 独立子任务
sequential 顺序执行 有依赖关系
map-reduce 映射归约 批量处理 + 聚合
pipeline 流水线 多阶段处理
fan-out 扇出探索 多方案对比

2. Production Workflows

预配置的生产级工作流:

  • Code Review - 自动代码审查
  • Test Gen - 测试用例生成
  • Doc Gen - 文档自动生成
  • CI/CD - 持续集成配置

3. Self-Improvement System

自进化系统组件:

  • WAL Protocol - 预写日志协议
  • Working Buffer - 工作缓冲区
  • Pattern Recognition - 模式识别

📁 文件结构

harness-engineering/
├── SKILL.md                  # 本文件
├── README.md                 # 项目文档
├── HARNESS-ARCHITECTURE.md   # 架构设计
├── WORKFLOWS.md              # 工作流指南
├── SELF-IMPROVEMENT.md       # 自进化文档
├── SECURITY-AUDIT.md         # 安全审计
├── harness/
│   ├── orchestrator.js       # 核心编排器
│   ├── patterns/             # 任务分解模式
│   └── utils/                # 工具模块
├── workflows/
│   ├── code-review.js        # 代码审查
│   ├── test-gen.js           # 测试生成
│   ├── doc-gen.js            # 文档生成
│   └── cicd.js               # CI/CD 集成
├── memory/
│   ├── wal.js                # WAL 协议
│   └── working-buffer.js     # 工作缓冲区
├── examples/
│   └── harness-demo.js       # 完整演示
└── tests/
    ├── test-suite.js         # 测试套件
    └── quick-test.js         # 快速验证

🔧 配置选项

Orchestrator 配置

const config = {
  maxParallel: 5,           // 最大并行数
  timeoutSeconds: 300,      // 超时时间
  retryAttempts: 2,         // 重试次数
  retryDelay: 1000,         // 重试延迟 (ms)
  enableLogging: true,      // 启用日志
};

Workflow 配置

const config = {
  // Code Review
  enableLint: true,
  enableSecurity: true,
  enablePerformance: true,
  enableTests: true,
  minApprovalScore: 0.8,
  
  // Test Gen
  framework: 'jest',
  minCoverage: 80,
  includeEdgeCases: true,
  
  // 通用
  autoCommit: false,
  autoComment: false,
};

📊 性能指标

根据基准测试:

组件 平均耗时 说明
Orchestrator ~50ms 5 个子任务并行
WAL ~2ms 单次事务
Buffer ~1ms 单次读写

测试通过率: 100% (23+ 测试用例)

🛡️ 安全状态

  • 总体评分: 82.5/100
  • 严重风险: 0
  • 高风险: 0
  • 中风险: 2 (待修复)

详见 SECURITY-AUDIT.md

📖 使用示例

示例 1: 并行代码审查

const result = await orchestrator.execute({
  task: '审查 PR #123',
  pattern: 'parallel',
  subTasks: [
    { task: '代码风格检查', agent: 'linter' },
    { task: '安全漏洞扫描', agent: 'security' },
    { task: '性能分析', agent: 'performance' },
  ]
});

示例 2: CI/CD 流水线

const result = await orchestrator.execute({
  task: '发布 v1.0.0',
  pattern: 'pipeline',
  stages: [
    { name: 'build', tasks: ['npm install', 'npm run build'] },
    { name: 'test', tasks: ['npm test', 'npm run lint'] },
    { name: 'deploy', tasks: ['deploy to prod'] },
  ]
});

示例 3: 多方案设计

const result = await orchestrator.execute({
  task: '设计认证系统',
  pattern: 'fan-out',
  subTasks: [
    { task: '方案 1: JWT', agent: 'architect' },
    { task: '方案 2: Session', agent: 'architect' },
    { task: '方案 3: OAuth', agent: 'architect' },
  ],
  fanIn: {
    task: '对比方案,推荐最佳',
    agent: 'chief-architect',
  }
});

🧪 测试

运行测试

node tests/test-suite.js

运行演示

node examples/harness-demo.js

📚 相关文档

🤝 贡献

  1. Fork 项目
  2. 创建特性分支
  3. 提交更改
  4. 推送到分支
  5. 创建 Pull Request

贡献代码时请在 .learnings/LEARNINGS.md 记录学习内容。

📄 许可证

MIT License

🙏 致谢

  • 多比 (Dobby) - 原始作者
  • OpenClaw 社区 - 平台支持

Skill 版本:1.0.0 | 最后更新:2026-04-18

安全使用建议
This skill appears to implement what it claims, but it persistently stores task state and logs in the skill workspace and includes scripts that push code to GitHub. Before installing or running it: - Review the WAL and WorkingBuffer code (memory/wal.js, memory/working-buffer.js) to confirm what data is logged and ensure no secrets (passwords, tokens, private keys) will be written. - Run the skill in an isolated/sandboxed environment (non‑root user, container, or VM) so logs and buffers are confined. - Set strict file permissions (chmod 600) on the skill's memory/wal and memory/buffer directories or enable disk encryption for those paths. - Disable or avoid features that auto‑commit, auto‑comment, or auto‑publish until you understand their behavior; keep autoCommit/autoComment=false by default. - Inspect push-to-github.sh before running. Do not paste your GitHub PAT into untrusted prompts/scripts; prefer authenticated git configured in a secure terminal session or use deploy keys scoped to a repository with minimal privileges. - If you need the self‑improvement features but are concerned about privacy, consider modifying the code to redact sensitive fields before writing logs or to enable encryption of WAL entries. If you want further analysis, provide the specific source files (memory/wal.js and memory/working-buffer.js) and I can point out exact lines to change to harden storage and logging.
功能分析
Type: OpenClaw Skill Name: dobby-harness Version: 1.0.0 The 'dobby-harness' skill is a comprehensive framework for multi-agent orchestration, providing tools for task decomposition, parallel execution, and state persistence. It implements a 'Self-Improvement System' using a Write-Ahead Logging (WAL) protocol (memory/wal.js) and a persistent working buffer (memory/working-buffer.js) to manage agent state and recovery. While the bundle includes a shell script (push-to-github.sh) with a hardcoded GitHub repository and performs extensive file system operations for logging, these behaviors are clearly aligned with the stated purpose of providing a robust, production-grade orchestration harness. The inclusion of a detailed self-security audit (SECURITY-AUDIT.md) that transparently identifies potential vulnerabilities (e.g., lack of log encryption) further supports the conclusion that the code is a legitimate tool rather than intentional malware.
能力标签
cryptorequires-oauth-tokenrequires-sensitive-credentials
能力评估
Purpose & Capability
The name/description match the included code and docs: orchestrator, patterns, workflows, WAL, and working buffer are present and implement the claimed capabilities. There are no unrelated required env vars or binaries in the manifest.
Instruction Scope
SKILL.md and other docs instruct running node examples and tests and include a publish guide and push script that operate on local filesystem paths (e.g., /home/admin/.openclaw/workspace) and call git. Those runtime instructions will read/write files under the skill workspace and may ask the user to run git push (which requires a PAT). No instructions were found that surreptitiously exfiltrate data to unknown remote endpoints, but the publish script and guides require manual credential use.
Install Mechanism
There is no automated install spec (instruction-only install), and all code is bundled with the skill. No remote downloads, package installs, or extract-from-URL steps were detected in the provided manifest — this reduces supply‑chain install risk.
Credentials
The skill declares no required environment variables or credentials. The only place credentials are referenced is the push-to-github.sh / publish guide which instruct the user how to obtain and use a GitHub PAT for pushing/publishing — a normal publishing step but one that requires caution. There are no unexpected credential requests in SKILL.md.
Persistence & Privilege
The skill persists runtime state via a WAL and a Working Buffer (memory/wal.js, memory/working-buffer.js) under its workspace. SECURITY-AUDIT.md flags these as storing plaintext logs and lacking access control. always:false and no cross-skill config edits are requested, but persistent local logs can contain sensitive task data and should be protected.
如何使用
  1. 确保已安装 OpenClaw(本地或 Docker 部署)
  2. 在对话框中输入安装命令:/install dobby-harness
  3. 安装完成后,直接呼叫该 Skill 的名称或使用 /dobby-harness 触发
  4. 根据 Skill 的参数说明提供必要输入,即可获得结构化输出
版本历史
v1.0.0
dobby-harness 1.0.0 – Initial Release - Introduces a production-ready multi-agent orchestration system with task decomposition, parallel execution, result aggregation, and self-improving capabilities. - Supports five orchestration patterns: parallel, sequential, map-reduce, pipeline, and fan-out. - Provides out-of-the-box production workflows including code review, test generation, documentation, and CI/CD integration. - Offers persistent state management via WAL protocol and working buffer modules. - Includes comprehensive examples, configuration options, and core documentation for quick onboarding. - Features documented test suite with full coverage and initial security audit report.
元数据
Slug dobby-harness
版本 1.0.0
许可证 MIT-0
累计安装 0
当前安装数 0
历史版本数 1
常见问题

Dobby Harness Self-improving Coding Skills 是什么?

Provides multi-agent orchestration with task decomposition, parallel execution, result aggregation, and self-improving workflows for complex coding tasks. 它是一个面向 Claude Code / OpenClaw 的 AI Agent Skill 插件,目前累计下载 60 次。

如何安装 Dobby Harness Self-improving Coding Skills?

在 OpenClaw 或 Claude Code 对话框中运行命令「/install dobby-harness」即可一键安装,无需额外配置。

Dobby Harness Self-improving Coding Skills 是免费的吗?

是的,Dobby Harness Self-improving Coding Skills 完全免费,采用 MIT-0 许可证,可自由下载、安装和使用。

Dobby Harness Self-improving Coding Skills 支持哪些平台?

Dobby Harness Self-improving Coding Skills 跨平台运行,可在任意部署了 OpenClaw / Claude Code 的环境中使用(cross-platform)。

谁开发了 Dobby Harness Self-improving Coding Skills?

由 Panb-KG(@panb-kg)开发并维护,当前版本 v1.0.0。

💬 留言讨论