Boss

Community

echoVic

boss

可审计的 agent 团队：BMAD 全自动研发流水线编排器。编排 9 个专业 Agent（PM、架构师、UI Designer、Tech Lead、Scrum Master、Frontend、Backend、QA、DevOps）从需求到部署，每一步都有事件溯源 + 不可绕过门禁 + 确定性 eval——可验证测试真跑、门禁真过。支持单环节切片命令（/boss:plan /review /qa /ship）与无 CLI 纯 Markdown 降级。 Triggers（全流水线）: 'boss mode', '/boss', '全自动开发', '从需求到部署', '帮我做一个', 'build this', 'ship it', '全流程', '自动化开发', '一键开发', 'start a project', 'new feature' Triggers（单环节切片，对存量项目最常用）: - /boss:review — '帮我评审', '代码评审', '技术评审', '看看这段代码有什么问题', '风险评估', 'review this', 'review my PR', 'code review' - /boss:qa — '跑一下测试', '补测试', '检查质量', '测试覆盖够不够', '门禁过了吗', 'run the tests', 'check quality', 'add tests' - /boss:plan — '先出个方案', '写个 PRD', '设计一下架构', 'write a spec', 'plan this feature' - /boss:ship — '可以发布了吗', '部署检查', 'ready to ship', 'deploy checks' Does NOT trigger: - 单文件修改或简单 bug 修复（直接编辑即可） - 纯代码阅读或解释（使用 read 工具） - 已有 pipeline 正在运行时的重复启动 - 极小事（预计 <30 分钟人工可完成、不需要 PRD/架构/门禁记录） Output: 完整项目代码 + PRD/架构/UI/测试/部署文档，写入 .boss/<feature>/ 目录

Overview

Publisher	echoVic
Repository	`boss-skill`
Skill name	`boss`
Stars	562
Forks	51
Bundled files	210
License	MIT
Links	Skill source Repository

Markdown instructions
A SKILL.md file the model loads on demand, so it only costs tokens when a request actually matches.
Works with any LLM
AI skills are plain Markdown, not provider-specific code, so this works with GPT, Claude, Gemini, Grok, or a local model.
210 bundled files
Scripts, templates, and references the model can read while it works. Files are read-only and never executed.
Open source
Published by echoVic on GitHub. Read the source before you install it.

Installation

Install the Boss AI skill in TypingMind to use it with any LLM, or drop it into another agent that reads SKILL.md.

Install in TypingMind

TypingMind installs a skill straight from its GitHub folder — it reads SKILL.md, bundles the resource files, and stores the result locally.

Open the app and go to Plugins → Skills.
Choose "Install from GitHub".
Paste the skill folder URL below and confirm.
https://github.com/echoVic/boss-skill/tree/main/skill
Enable the skill in any chat where you want it available.

Plugins → Skills → Add skill → From GitHub URL, then paste the folder URL and press Continue.

Install in another agent

Any agent that reads the Agent Skills format can use this skill — copy the folder into that agent's skills directory.

Claude Code — .claude/skills

git clone --depth 1 https://github.com/echoVic/boss-skill.git /tmp/boss-skill
mkdir -p .claude/skills
cp -r /tmp/boss-skill/skill .claude/skills/boss

Restart Claude Code after copying so it picks up the new skill.

Use it in TypingMind

Enable Boss in any TypingMind chat and the model takes it from there. Its name and description sit in the system prompt, and the moment a request matches, the model loads the full instructions itself — you never invoke it by hand, and it costs no tokens until it is actually used.

The model loads Boss on its own as soon as a request matches it.

Works with any AI model

AI skills are plain Markdown instructions rather than provider-specific code, so Boss is not tied to the model it was written for. Install it once in TypingMind and use it with GPT-5, Claude, Gemini, Grok, DeepSeek, Mistral, Llama, or a local model you run yourself — all on your own API keys.

Loaded only when it is needed
The system prompt carries just the name and description. The instructions are fetched on the first matching request, so an idle skill costs nothing.
Switch models mid-chat
Because the skill is instructions rather than code, changing model does not break it — the next model reads the same SKILL.md.

Skill instructions

This is the SKILL.md content the model loads. Read it before installing — a skill is instructions your model will follow.

Boss - BMAD Harness Orchestrator

你是 Boss Agent，负责编排完整软件开发生命周期。不要把本文件当百科全书；本文件只保存入口、不变量和按需读取索引。

不变量

你负责编排，不直接替专业 Agent 写正式产物。
状态真相源是 runtime 事件流。 不直接编辑 .boss/<feature>/.meta/execution.json；所有状态变更通过 boss runtime ...。
产物驱动。 文档和代码产物写入 .boss/<feature>/，完成后用 boss runtime record-artifact 记录。
质量门禁不可绕过。 测试、Wave 边界校验、最终 gate 失败时不得宣布交付完成。
渐进式披露。 只读取当前步骤需要的 reference、Agent prompt、template；不要一次性加载所有协议全文。
平台适配不改变状态语义。 Claude Code、Codex、OpenClaw、Antigravity、Hermes 都以 Runtime Core 为准。
中文交付。 Boss 生成的 .boss/<feature>/ 文档默认使用中文。
不重造宿主已有的能力。 Boss 只提供宿主没有的东西：事件溯源审计、不可绕过的门禁、产物 provenance。派发子 Agent、读写文件、检索代码、会话内任务清单一律用宿主原语。

职责边界：Boss 提供什么、宿主提供什么

能力	归属	说明
子 Agent 派发	宿主	用宿主的 Task / Agent 工具，Boss 不自建调度器
文件读写、代码检索	宿主	Read / Write / Edit / Glob / Grep
会话内临时清单	宿主	宿主的 todo 原语；Boss 的 todo 仅指会话收敛产出的
带 owner 与 successCriteria 的跨 Agent 交付项，落在事件流中
结构化返回	宿主 + 工具层校验	状态经 `report-agent-status` 校验枚举，不解析散文
事件溯源与重放	Boss	`events.jsonl` 追加写 + projector 重放，宿主不提供
不可绕过门禁	Boss	建模为 runtime stage，非提示词自觉
产物 provenance	Boss	对能改变 Agent 行为的文件做内容指纹
记忆 / 偏好派生	Boss，确定性	从 `events.jsonl` 规则化投影（含用户选择偏好聚合），
可完整重放；不外挂第二个 LLM 去二次抽取

什么时候读取什么

场景	读取
进入 `/boss` 或 boss mode，准备跑完整流水线	`references/orchestration-loop.md`
需要 runtime 命令、workflow-plan、resume、事件真相源	`references/runtime-surface.md`
准备派发 code Agent、检查 Evidence Wave、Blast Radius、写集冲突	`references/evidence-waves.md`
需要解释或适配 Claude/Codex/OpenClaw/Antigravity/Hermes 行为	`references/platform-drivers.md`
调试 hooks、session resume/end、artifact guard	`references/hooks-runtime.md`
准备或记录文档/JSON 产物	`references/artifact-guide.md`
需要质量门禁细节	`references/quality-gate.md`
需要测试证据要求	`references/testing-standards.md`
需要 BMAD 背景或历史设计说明	`references/bmad-methodology.md`
未检测到 `boss` CLI，需要纯 Markdown 降级运行	`references/no-cli-fallback.md`
需要扩展 Boss（自定义 Agent / pack / gate 插件）	`references/extending-boss.md`
派发子 Agent 前建立公共协议 prefix 缓存	`agents/shared/protocol-manifest.md`
子 Agent 状态格式、会话原语、REVISION_NEEDED	`agents/prompts/subagent-protocol.md`
需要技术栈探测协议	`agents/shared/tech-detection.md`

快速入口

判断用户输入是否是可执行任务。约束类输入不要新建 .boss/<feature>/；按 references/orchestration-loop.md 的 Feature Slug 归一化处理。
CLI 探测：派发前运行 boss --version。CLI 随本 skill 分发（市场安装自带，无需 npm）：
- 插件安装（Claude Code / Codex）：node "${CLAUDE_PLUGIN_ROOT}/skill/cli/bin/boss.mts" --version
- 复制安装 / skills.sh：node <skill>/cli/bin/boss.mts --version（<skill> 即包含本 SKILL.md 的目录；需 Node >=22.18）
- PATH 上已有 boss 时可直接用 boss 失败则进入降级模式，读取 references/no-cli-fallback.md，用 .boss/<feature>/STATE.md 承载状态（CLI 是可审计性增强，不是准入条件）。下文所有 boss <command> 均指运行上述入口。
除非 --quick，先确认”做什么 + 给谁用 + 核心场景”。缺失时读取 skills/brainstorming/SKILL.md。
初始化或恢复项目：
- 新 feature：boss project init <feature-name>
- 低阶 runtime 初始化：boss runtime init-pipeline <feature>
- 继续执行图：boss runtime resume <feature> --from-run <run-id>
调用 boss packs detect <project-dir> --json，再根据 ready artifacts 循环派发 Agent。
每个产物完成后调用 boss runtime record-artifact <feature> <artifact-name> <N>。
code 阶段前必须读取 references/evidence-waves.md；缺 Repo Preflight、Evidence Wave、Contract Matrix、写集 owner、Blast Radius 任一项，不得派发 code Agent。
收尾前运行适用测试、门禁和 boss runtime generate-summary <feature>。

Agent 路由

Agent	文件	读取时机
PM	`agents/boss-pm.md`	生成或修订 PRD
Architect	`agents/boss-architect.md`	架构设计、API/数据模型、技术选型
UI Designer	`agents/boss-ui-designer.md`	`ui-spec.md` 与 `ui-design.json`
Tech Lead	`agents/boss-tech-lead.md`	技术评审、风险评估
Scrum Master	`agents/boss-scrum-master.md`	`tasks.md`、Evidence Wave、写集规划
Frontend	`agents/boss-frontend.md`	前端实现，优先遵循 `ui-design.json`
Backend	`agents/boss-backend.md`	API、数据库、后端测试
QA	`agents/boss-qa.md`	核心用户路径、真实 payload、越权/分页/旧数据验证
DevOps	`agents/boss-devops.md`	构建、部署、健康检查

子代理协议

先读取 agents/shared/protocol-manifest.md，使用协议 manifest、prefix 缓存、按需加载和渐进式披露。
只有状态格式或会话升级不清楚时，再读取 agents/shared/agent-protocol.md 或 agents/prompts/subagent-protocol.md。
子代理状态只接受 DONE / DONE_WITH_CONCERNS / NEEDS_CONTEXT / BLOCKED / REVISION_NEEDED。
子代理自报 DONE 不等于验收通过；必须跑 Wave 边界校验或对应门禁。

命令入口地图

/boss 是完整流水线；下列切片命令是同一底层的单环节独立入口，复用相同 agent prompt、runtime 与 .boss/<feature>/ 产物，适合只想用其中一环或对已有项目单点介入的场景。

命令	环节	底层
`/boss`	完整 4 阶段流水线	`references/orchestration-loop.md`
`/boss:plan`	规划（PM + Architect）	`agents/boss-pm.md`、`agents/boss-architect.md`
`/boss:review`	技术评审（Tech Lead，只读）	`agents/boss-tech-lead.md`
`/boss:qa`	测试 + 门禁（QA）	`agents/boss-qa.md`、`references/quality-gate.md`
`/boss:ship`	构建部署（DevOps + Gate 2）	`agents/boss-devops.md`
`/boss:extend`	引导式扩展自定义 agent / pack	`references/extending-boss.md`

语言参数

所有命令支持 --lang <zh|en> 控制 .boss/<feature>/ 产物语言。未显式指定时默认 zh（不变量 7）。切片命令与 /boss 共享该参数语义。

项目扩展点

项目级模板：.boss/templates/
项目级插件：.boss/plugins/
项目级 pipeline packs：.boss/pipeline-packs/
项目级 Artifact DAG：.boss/artifact-dag.json
可渲染 UI 设计产物：.boss/<feature>/ui-design.json，预览命令 boss design preview <feature>

输出格式

最终回答包含：

feature 名称和 .boss/<feature>/ 产物路径
关键文档路径
测试和门禁摘要
未完成事项或阻塞原因
部署/预览 URL（如有）

Bundled files

The model reads these on demand while the skill is loaded. They are exposed as readable files and are never executed.

and 140 more files.

Frequently asked questions

What does the Boss AI skill do?

Why use Boss on TypingMind?

Because you install it once and use it with any model. Boss is plain Markdown rather than provider-specific code, so the same skill runs on GPT-5, Claude, Gemini, Grok, or a local model — and you can switch model mid-chat without it breaking. TypingMind runs on your own API keys, so you pay providers directly instead of a per-seat subscription, and your skills and chats stay in your own storage.

How do I install Boss in TypingMind?

Open Plugins → Skills → Install from GitHub in TypingMind and paste https://github.com/echoVic/boss-skill/tree/main/skill. TypingMind reads its SKILL.md and bundles its files and installs it as a skill you can enable per chat.

Which AI models can use Boss?

Any model you connect in TypingMind. AI skills are plain Markdown instructions rather than provider-specific code, so GPT, Claude, Gemini, Grok, and local models can all load this skill when a request matches it.

How many AI models can I use with Boss?

As many as you like. As long as a model supports skills, you can use Boss with it — GPT, Claude, Gemini, Grok, DeepSeek, Mistral, Llama and more — all on TypingMind with your own API keys.

Is the Boss AI skill free?

Yes. It is published on GitHub by echoVic under the MIT license. You only pay your own AI provider for the tokens you use.

What are AI skills?

An AI skill is a reusable instruction bundle that teaches an AI model how to do one specific task. It follows the open Agent Skills format: a SKILL.md file with a name and description, plus any scripts, templates or reference files the model may need. The model reads the instructions only when your request matches the skill, so an installed skill costs nothing until it is used.

How are AI skills different from plugins or MCP servers?

A plugin or MCP server gives a model new tools to call — code that runs somewhere and returns a result. An AI skill gives the model knowledge and process instead: how to approach a task, which steps to follow, what good output looks like. Skills are plain Markdown, so they need no server, no API key and no runtime, and they work with any model.

View all

Clawhub

openclaw

OrganizationPopular

Search ClawHub for skills when a requested capability is not already available; install, verify, update, uninstall, publish, or sync skills.

Diagnosing Superpowers

obra

CommunityPopular

Use when a superpowers session went wrong and your human partner wants to know why — repeated work, ignored plans, stumbles, poor results, a skill that didn't fire, "it took too long", "why is it so…

Brainstorming

You MUST use this before any creative work - creating features, building components, adding functionality, or modifying behavior. Explores user intent, requirements and design before implementation.

Executing Plans

Use when you have a written implementation plan to execute in a separate session with review checkpoints

Overview

Installation

Install in TypingMind

Install in another agent

Use it in TypingMind

Works with any AI model

Skill instructions

Boss - BMAD Harness Orchestrator

不变量

职责边界：Boss 提供什么、宿主提供什么

什么时候读取什么

快速入口

Agent 路由

子代理协议

命令入口地图

语言参数

项目扩展点

输出格式

Bundled files

Frequently asked questions

Related AI Skills

Set up your own AI workspace now