By derekhu0002
Complete collection of battle-tested Claude Code configs from an Anthropic hackathon winner - agents, skills, hooks, and rules evolved over 10+ months of intensive daily use
Sync documentation with the codebase, generating from source-of-truth files.
Answer a quick side question without interrupting or losing context from the current task. Resume work automatically after answering.
Incrementally fix build and type errors with minimal, safe changes.
Create or verify a checkpoint in your workflow.
Start NanoClaw v2 — ECC's persistent, zero-dependency REPL with model routing, skill hot-load, branching, compaction, export, and metrics.
Software architecture specialist for system design, scalability, and technical decision-making. Use PROACTIVELY when planning new features, refactoring large systems, or making architectural decisions.
Build and TypeScript error resolution specialist. Use PROACTIVELY when build fails or type errors occur. Fixes build/type errors only with minimal diffs, no architectural edits. Focuses on getting the build green quickly.
Personal communication chief of staff that triages email, Slack, LINE, and Messenger. Classifies messages into 4 tiers (skip/info_only/meeting_info/action_required), generates draft replies, and enforces post-send follow-through via hooks. Use when managing multi-channel communication workflows.
Expert code review specialist. Proactively reviews code for quality, security, and maintainability. Use immediately after writing or modifying code. MUST BE USED for all code changes.
C++ build, CMake, and compilation error resolution specialist. Fixes build errors, linker issues, and template errors with minimal changes. Use when C++ builds fail.
Design and optimize AI agent action spaces, tool definitions, and observation formatting for higher completion rates.
Operate as an agentic engineer using eval-first execution, decomposition, and cost-aware model routing.
Engineering operating model for teams where AI agents generate a large share of implementation output.
Regression testing strategies for AI-assisted development. Sandbox-mode API testing without database dependencies, automated bug-check workflows, and patterns to catch AI blind spots where the same model writes and reviews code.
SCRUM flow skill for review and quality inspection of design-code alignment.
Matches all tools
Hooks run on every tool call, not just specific ones
Executes bash commands
Hook triggers when Bash tool is used
Own this plugin?
Verify ownership to unlock analytics, metadata editing, and a verified badge. GitHub access is read-only (username + org membership).
Sign in to claimOwn this plugin?
Verify ownership to unlock analytics, metadata editing, and a verified badge. GitHub access is read-only (username + org membership).
Sign in to claimBased on adoption, maintenance, documentation, and repository signals. Not a security audit or endorsement.
Modifies files
Hook triggers on file write and edit operations
Modifies files
Hook triggers on file write and edit operations
Uses power tools
Uses Bash, Write, or Edit tools
Uses power tools
Uses Bash, Write, or Edit tools
No model invocation
Executes directly as bash, bypassing the AI model
No model invocation
Executes directly as bash, bypassing the AI model
Runs pre-commands
Contains inline bash commands via ! syntax
Runs pre-commands
Contains inline bash commands via ! syntax
Bash prerequisite issue
Uses bash pre-commands but Bash not in allowed tools
Bash prerequisite issue
Uses bash pre-commands but Bash not in allowed tools
AI4PB Orchestrator 是一个 VS Code 扩展,用来把企业架构模型、AI 编码执行、设计审计和迭代复盘串成一条可重复的交付流程。
它面向的不是“随便让 AI 帮我写点代码”的场景,而是“让 AI 在架构约束下参与真实项目交付”的场景。
很多团队开始用 AI 写代码后,会很快遇到几个问题:
AI4PB 的解决方式是:
一句话概括:
不是让 AI 自由发挥,而是让 AI 在你的架构和流程里工作。
AI4PB 把工作分成两端:
.aicodingconfig 中按需切换,详见下方"快速开始"章节)核心闭环如下:
design/KG/SystemArchitecture.jsondesign/tasks/taskandissues_for_LLM.md)npm install -g opencode-aiopencode serve(默认监听 http://127.0.0.1:4096).aicodingconfig 中配置地址与路由策略,示例如下:{
"AGENT_ROUTER_CONFIG": {
"default_agent": "opencode",
"opencode": {
"transport": "server",
"executionHost": "wsl",
"server": {
"baseUrl": "http://127.0.0.1:4096"
}
}
}
}
如果你是使用者:
如果你是维护者,在仓库根目录执行:
npm install
npm run compile
npm run release:vsix:nobump
打开 VS Code 后:
在架构元素上挂载 Task / Issue,建议遵循:
Status = ActiveAssigned To = llmName 写成可执行任务说明然后从 EA 导出:
design/KG/SystemArchitecture.json任务清单(design/tasks/taskandissues_for_LLM.md)由扩展在执行 提取任务 步骤时自动生成,无需从 EA 手动导出。
当前 AI4PB 侧边栏中的主操作按钮如下,按钮名称与界面保持一致:
| 区域 | 当前侧边栏动作名称 | 说明 |
|---|---|---|
| 环境配置 | EA模板初始化 | 初始化 EA 模板与基础工作区结构 |
| 环境配置 | 初始化AICodingAgent配置 | 将扩展内置的 .github、.opencode、skills、.agents、.claude、rules 初始化到当前工作区 |
| 环境配置 | EA导出参数配置 | 配置 .aicodingconfig / .aicodingconfig.json 中的导出与维护参数 |
| 环境配置 | 参数查询 | 查看当前导出参数摘要 |
| 计划与汇报流程 | 提取任务 | 从 design/KG/SystemArchitecture.json 中提取维护任务并生成任务反馈输入 |
| 计划与汇报流程 | 待办梳理 | 生成当前迭代待办清单 |
| 计划与汇报流程 | 周报输出 | 生成管理层周报 |
| 执行支持流程 | 执行支持 | 为任务生成执行说明并进入实现支持 |
| SCRUM敏捷开发流程 | 待办梳理 | 回看当前 Sprint 的优先级与待办范围 |
| SCRUM敏捷开发流程 | 迭代启动 | 为本轮迭代建立上下文、范围和目标 |
| SCRUM敏捷开发流程 | 问题反馈 | 打开问题反馈编辑页,回填 ResolverNotes 并确认写入反馈 |
| SCRUM敏捷开发流程 | 问题处理 | 根据 Issue 记录继续多轮缺陷修复 |
| SCRUM敏捷开发流程 | 迭代收尾 | 对本轮交付范围做收尾检查 |
| SCRUM敏捷开发流程 | 提交总结 | 生成提交总结与 Git Commit Message |
| 设计审计流程 | 设计审计 | 执行代码与架构对齐审计并输出 design/temp/audit.md |
切换 AI 引擎(Copilot / OpenCode)
扩展默认使用 GitHub Copilot。若需切换为 OpenCode,在项目根目录 .aicodingconfig 中修改:
{
"AGENT_ROUTER_CONFIG": {
"default_agent": "opencode"
}
}
也可仅对特定阶段路由到 OpenCode,其余阶段保持 Copilot:
{
"AGENT_ROUTER_CONFIG": {
"default_agent": "copilot",
"task_specific_agents": {
"task-list": "opencode"
}
}
}
AI4PB 并不是只“调用一个外部 OpenCode 命令”这么简单。扩展本身内置了一套 .opencode 资产包,用于把 OpenCode 作为可初始化、可复制、可治理的执行环境接入工作区。
这套内置资产至少包括:
.opencode/package.json、.opencode/index.ts).opencode/plugins/*).opencode/tools/*),包含 run-tests、check-coverage、security-audit、format-code、lint-check、git-summary.opencode/skills/*、.opencode/opencode.json)扩展当前支持两种 OpenCode 传输模式,统一通过 .aicodingconfig 的 AGENT_ROUTER_CONFIG.opencode 配置:
transport = "cli":直接调用本机或 WSL 中的 opencode 命令,适合本地命令式执行transport = "server":通过 opencode serve 暴露的 HTTP 服务进行会话创建、事件订阅与流式响应,适合常驻服务模式此外,扩展会根据 executionHost、wslDistribution、server.directory、server.workspace 等配置自动处理 Windows 与 WSL 路径归一化,保证 CLI 模式和 Server 模式都能在同一工作区路由规则下工作。
安装扩展后,VS Code 左侧活动栏出现 AI4PB DEV 图标,点击后打开工作流面板。
面板按 SCRUM 阶段排列,提供从架构初始化到迭代复盘的全套操作入口:
opencode serve 的 Server 模式;两种模式都会由扩展统一接管并在面板内展示执行反馈不需要手动复制 Prompt,不需要记忆命令,每个迭代阶段都有一键操作入口。
业务交付物:
implementation/task-list.mdimplementation/taskhelpinfos/*.mdimplementation/reports/*.md过程工件:
design/temp/audit.mddebug/iteration-commit-message.mdnpx claudepluginhub derekhu0002/ai4pb-orchestratorComplete collection of battle-tested Claude Code configs from an Anthropic hackathon winner - agents, skills, hooks, and rules evolved over 10+ months of intensive daily use
AI team workflow harness for Claude Code — specialist agents, shared skills, slash commands, hooks, and staged delegation pipeline.
Elite AI development framework: reference-first design, agent orchestration, automated quality gates, and battle-tested engineering workflows
Harness for Claude Code — skills, /harness:* slash commands, persona subagents, lifecycle hooks, and MCP tools without per-repo `harness setup`. Sibling plugins exist for Cursor, Gemini CLI, and Codex.
CodeStable AI coding workflow skills.
13 agents, 35 skills, 19 commands, 15 hooks — spec-driven, multi-agent orchestration for Claude Code with per-phase model tiers, opt-in lifecycle hooks, and optional cross-device semantic memory.