setup — for Claude Code clawBro, community, for Claude Code, ide skills, clawbro-rust-agent, add-lark, add-dingtalk, add-acp-backend, add-agent, add-team-mode

v1.0.0

Acerca de este Skill

Escenario recomendado: Ideal for AI agents that need clawbro gateway 初始化引导. Resumen localizado: 🦀 Let CLI Coding Agents Work Like OpenClaw in Chat and collaborating as a team at all times. This AI agent skill supports Claude Code, Cursor, and Windsurf workflows.

Características

ClawBro Gateway 初始化引导
本 Skill 通过对话引导你完成 /.clawbro/config.toml 的创建和配置。
所有变更均为配置文件修改, 不需要 fork 项目,不需要重新编译 。
默认使用内置的 clawbro-rust-agent 作为 AI 执行核心,无需安装任何额外 CLI 工具。
/add-lark — 添加飞书(Lark)Channel

# Core Topics

FISHers6 FISHers6
[31]
[7]
Updated: 4/10/2026

Killer-Skills Review

Decision support comes first. Repository text comes second.

Reference-Only Page Review Score: 10/11

This page remains useful for teams, but Killer-Skills treats it as reference material instead of a primary organic landing page.

Original recommendation layer Concrete use-case guidance Explicit limitations and caution Quality floor passed for review
Review Score
10/11
Quality Score
57
Canonical Locale
zh
Detected Body Locale
zh

Escenario recomendado: Ideal for AI agents that need clawbro gateway 初始化引导. Resumen localizado: 🦀 Let CLI Coding Agents Work Like OpenClaw in Chat and collaborating as a team at all times. This AI agent skill supports Claude Code, Cursor, and Windsurf workflows.

¿Por qué usar esta habilidad?

Recomendacion: setup helps agents clawbro gateway 初始化引导. 🦀 Let CLI Coding Agents Work Like OpenClaw in Chat and collaborating as a team at all times. This AI agent skill supports Claude Code, Cursor, and Windsurf

Mejor para

Escenario recomendado: Ideal for AI agents that need clawbro gateway 初始化引导.

Casos de uso accionables for setup

Caso de uso: Applying ClawBro Gateway 初始化引导
Caso de uso: Applying 本 Skill 通过对话引导你完成 /.clawbro/config.toml 的创建和配置。
Caso de uso: Applying 所有变更均为配置文件修改, 不需要 fork 项目,不需要重新编译 。

! Seguridad y limitaciones

  • Limitacion: Requires repository-specific context from the skill documentation
  • Limitacion: Works best when the underlying tools and dependencies are already configured

Why this page is reference-only

  • - Current locale does not satisfy the locale-governance contract.

Source Boundary

The section below is imported from the upstream repository and should be treated as secondary evidence. Use the Killer-Skills review above as the primary layer for fit, risk, and installation decisions.

After The Review

Decide The Next Action Before You Keep Reading Repository Material

Killer-Skills should not stop at opening repository instructions. It should help you decide whether to install this skill, when to cross-check against trusted collections, and when to move into workflow rollout.

Labs Demo

Browser Sandbox Environment

⚡️ Ready to unleash?

Experience this Agent in a zero-setup browser environment powered by WebContainers. No installation required.

Boot Container Sandbox

FAQ & Installation Steps

These questions and steps mirror the structured data on this page for better search understanding.

? Frequently Asked Questions

What is setup?

Escenario recomendado: Ideal for AI agents that need clawbro gateway 初始化引导. Resumen localizado: 🦀 Let CLI Coding Agents Work Like OpenClaw in Chat and collaborating as a team at all times. This AI agent skill supports Claude Code, Cursor, and Windsurf workflows.

How do I install setup?

Run the command: npx killer-skills add FISHers6/clawBro/setup. It works with Cursor, Windsurf, VS Code, Claude Code, and 19+ other IDEs.

What are the use cases for setup?

Key use cases include: Caso de uso: Applying ClawBro Gateway 初始化引导, Caso de uso: Applying 本 Skill 通过对话引导你完成 /.clawbro/config.toml 的创建和配置。, Caso de uso: Applying 所有变更均为配置文件修改, 不需要 fork 项目,不需要重新编译 。.

Which IDEs are compatible with setup?

This skill is compatible with Cursor, Windsurf, VS Code, Trae, Claude Code, OpenClaw, Aider, Codex, OpenCode, Goose, Cline, Roo Code, Kiro, Augment Code, Continue, GitHub Copilot, Sourcegraph Cody, and Amazon Q Developer. Use the Killer-Skills CLI for universal one-command installation.

Are there any limitations for setup?

Limitacion: Requires repository-specific context from the skill documentation. Limitacion: Works best when the underlying tools and dependencies are already configured.

How To Install

  1. 1. Open your terminal

    Open the terminal or command line in your project directory.

  2. 2. Run the install command

    Run: npx killer-skills add FISHers6/clawBro/setup. The CLI will automatically detect your IDE or AI agent and configure the skill.

  3. 3. Start using the skill

    The skill is now active. Your AI agent can use setup immediately in the current project.

! Reference-Only Mode

This page remains useful for installation and reference, but Killer-Skills no longer treats it as a primary indexable landing page. Read the review above before relying on the upstream repository instructions.

Upstream Repository Material

The section below is imported from the upstream repository and should be treated as secondary evidence. Use the Killer-Skills review above as the primary layer for fit, risk, and installation decisions.

Upstream Source

setup

🦀 Let CLI Coding Agents Work Like OpenClaw in Chat and collaborating as a team at all times. This AI agent skill supports Claude Code, Cursor, and Windsurf

SKILL.md
Readonly
Upstream Repository Material
The section below is imported from the upstream repository and should be treated as secondary evidence. Use the Killer-Skills review above as the primary layer for fit, risk, and installation decisions.
Supporting Evidence

ClawBro Gateway 初始化引导

关于本 Skill

本 Skill 通过对话引导你完成 ~/.clawbro/config.toml 的创建和配置。 所有变更均为配置文件修改,不需要 fork 项目,不需要重新编译

默认使用内置的 clawbro-rust-agent 作为 AI 执行核心,无需安装任何额外 CLI 工具。

其他扩展 Skill:

  • /add-lark — 添加飞书(Lark)Channel
  • /add-dingtalk — 添加钉钉(DingTalk)Channel
  • /add-acp-backend — 添加外部 ACP Agent(claude-code / codex / qwen 等)
  • /add-agent — 向 agent_roster 追加新 Agent
  • /add-team-mode — 为指定群组配置 Team 模式
  • /doctor — 诊断和修复配置问题

Phase 0:Pre-flight 检查

在开始前,先确认运行环境是否就绪。

0.1 检查 Binary 是否存在

运行以下命令,确认两个 binary 可以找到:

bash
1which clawbro-gateway || ls ./target/release/clawbro-gateway 2>/dev/null || ls ./target/debug/clawbro-gateway 2>/dev/null 2which clawbro-rust-agent || ls ./target/release/clawbro-rust-agent 2>/dev/null || ls ./target/debug/clawbro-rust-agent 2>/dev/null

如果找不到 binary,需要先编译:

bash
1cargo build -p clawbro-server --bin clawbro-gateway 2cargo build -p clawbro-rust-agent

编译完成后,建议把两个 binary 复制到 PATH:

bash
1cp target/debug/clawbro-gateway ~/.local/bin/ 2cp target/debug/clawbro-rust-agent ~/.local/bin/ 3# 确认 ~/.local/bin 在 PATH 中: 4echo $PATH | tr ':' '\n' | grep -q "$HOME/.local/bin" && echo "✓ PATH OK" || echo "⚠ 请在 shell profile 中添加: export PATH=\"\$HOME/.local/bin:\$PATH\""

0.2 创建运行时目录

bash
1mkdir -p ~/.clawbro/{sessions,shared,skills,personas} 2echo "✓ 目录已创建"

0.3 检查是否已有配置

bash
1[ -f ~/.clawbro/config.toml ] && echo "⚠ 已存在 config.toml,本次将覆盖" || echo "✓ 干净环境,将创建新配置"

如果已有配置且只想修改部分内容,请直接告诉我要修改什么,我会定点更新。


Phase 1:API Key 配置

clawbro-rust-agent 启动时从环境变量读取 API Key。

1.1 询问用户使用哪个 Provider

请告诉我你想使用哪个 AI Provider:

选项Provider环境变量推荐模型
1Anthropic(Claude)ANTHROPIC_API_KEYclaude-sonnet-4-6
2OpenAI(GPT)OPENAI_API_KEYgpt-4o
3DeepSeekOPENAI_API_KEY + OPENAI_API_BASEdeepseek-chat
4其他 OpenAI 兼容端点OPENAI_API_KEY + OPENAI_API_BASE按服务商

1.2 收集 API Key

请提供你的 API Key(输入后我会写入配置,不会在任何地方展示原文)。

1.3 写入环境变量

将 API Key 写入 ~/.clawbro/.env(gateway 启动时可 source 这个文件):

根据用户选择,写入对应内容:

Anthropic

bash
1cat > ~/.clawbro/.env << 'EOF' 2export ANTHROPIC_API_KEY=<用户填写的key> 3EOF

OpenAI

bash
1cat > ~/.clawbro/.env << 'EOF' 2export OPENAI_API_KEY=<用户填写的key> 3EOF

DeepSeek

bash
1cat > ~/.clawbro/.env << 'EOF' 2export OPENAI_API_KEY=<用户填写的key> 3export OPENAI_API_BASE=https://api.deepseek.com 4export CLAWBRO_MODEL=deepseek-chat 5EOF

其他 OpenAI 兼容

bash
1cat > ~/.clawbro/.env << 'EOF' 2export OPENAI_API_KEY=<用户填写的key> 3export OPENAI_API_BASE=<用户填写的base url> 4export CLAWBRO_MODEL=<用户填写的model> 5EOF

1.4 提示 shell 集成(可选)

询问用户:是否要把 source ~/.clawbro/.env 加到 shell profile?

如果同意,检测 shell 并追加:

bash
1SHELL_PROFILE="" 2case "$SHELL" in 3 */zsh) SHELL_PROFILE="$HOME/.zshrc" ;; 4 */bash) SHELL_PROFILE="$HOME/.bashrc" ;; 5esac 6if [ -n "$SHELL_PROFILE" ]; then 7 grep -q "clawbro/.env" "$SHELL_PROFILE" 2>/dev/null \ 8 && echo "✓ 已存在,跳过" \ 9 || echo 'source ~/.clawbro/.env' >> "$SHELL_PROFILE" && echo "✓ 已写入 $SHELL_PROFILE" 10fi

Phase 2:选择运行模式

询问用户要配置哪种模式:

请选择你的使用场景:

1. Solo(最简单)
   - 一个 AI Agent
   - 适合个人使用、快速上手
   - WebSocket 访问

2. Multi-agent(多 Agent 切换)
   - 多个命名 Agent,通过 @mention 切换
   - 不同 Agent 可以有不同 persona、workspace、backend
   - 适合区分"代码助手"/"写作助手"/"研究助手"

3. Team(多 Agent 协作)
   - Lead Agent 负责拆解任务、验收
   - Specialist Agent 并行执行子任务
   - 适合复杂工程任务、需要多专家协作的场景
   - 需要先配好多 Agent roster,再开启 Team 模式

建议顺序:先跑通 Solo,再升级到 Multi-agent,最后按需开 Team。

根据用户选择进入对应配置分支。


Phase 3A:Solo 模式配置

3A.1 收集基本参数

询问:

  • 工作目录(默认可空,Agent 将在当前目录工作):/Users/xxx/work
  • WebSocket Token(用于 /ws 端点鉴权,可留空 = 开放模式)
  • 监听端口(默认 8080,0 = 随机端口)

3A.2 生成 config.toml

写入 ~/.clawbro/config.toml

toml
1[gateway] 2host = "127.0.0.1" 3port = <用户选择的端口> 4require_mention_in_groups = false 5<如果有 default_workspace>default_workspace = "<用户填写的目录>" 6 7[auth] 8<如果有 ws_token>ws_token = "<用户填写的 token>" 9 10[agent] 11backend_id = "native-main" 12 13[[backend]] 14id = "native-main" 15family = "quick_ai_native" 16 17[backend.launch] 18type = "bundled_command" 19 20[session] 21dir = "/Users/<username>/.clawbro/sessions" 22 23[memory] 24shared_dir = "/Users/<username>/.clawbro/shared" 25distill_every_n = 20 26distiller_binary = "clawbro-rust-agent" 27 28[skills] 29dir = "/Users/<username>/.clawbro/skills"

Phase 3B:Multi-agent 模式配置

3B.1 收集 Agent 信息

询问需要几个 Agent,并对每个 Agent 收集:

  • Agent 名称(如 claudecodexresearcher
  • 触发 mention(如 @claude、@代码助手)
  • Backend(先都用 native-main,后续可用 /add-acp-backend 添加其他)
  • Workspace 目录(可选)
  • Persona 目录(可选,存放 SOUL.md/IDENTITY.md 等)

3B.2 是否需要 Binding(默认路由)

询问:某个 scope(如 Lark 群)默认走哪个 Agent?

如果需要,收集:

  • scope 格式(如 group:lark:chat-xxx
  • 默认 Agent 名称

3B.3 生成 config.toml

toml
1[gateway] 2host = "127.0.0.1" 3port = <端口> 4require_mention_in_groups = true 5 6[auth] 7<如有>ws_token = "<token>" 8 9[[backend]] 10id = "native-main" 11family = "quick_ai_native" 12 13[backend.launch] 14type = "bundled_command" 15 16[[agent_roster]] 17name = "<agent1_name>" 18mentions = ["<@mention1>"] 19backend_id = "native-main" 20<如有>persona_dir = "<路径>" 21<如有>workspace_dir = "<路径>" 22 23[[agent_roster]] 24name = "<agent2_name>" 25mentions = ["<@mention2>"] 26backend_id = "native-main" 27<如有>persona_dir = "<路径>" 28<如有>workspace_dir = "<路径>" 29 30<如有 binding> 31[[binding]] 32kind = "scope" 33agent = "<默认agent名>" 34scope = "<scope>" 35channel = "<channel>" 36 37[session] 38dir = "/Users/<username>/.clawbro/sessions" 39 40[memory] 41shared_dir = "/Users/<username>/.clawbro/shared" 42distill_every_n = 20 43distiller_binary = "clawbro-rust-agent" 44 45[skills] 46dir = "/Users/<username>/.clawbro/skills"

Phase 3C:Team 模式配置

Team 模式在 Multi-agent 基础上增加编排配置。

3C.1 先完成 Multi-agent 配置(3B 步骤)

至少需要:

  • 1 个 Lead Agent(front_bot
  • 1+ 个 Specialist Agent(roster

3C.2 选择 Team 作用范围

询问:Team 模式绑定到哪里?

选项 A:群组(Lark/DingTalk group)

  • 需要 [[group]] + scope = "group:lark:chat-xxx"
  • 群内消息驱动 Lead,Lead 分配给 Specialists

选项 B:单聊个人工作台(DM scope)

  • 需要 [[team_scope]] + 精确 scope(如 user:ou_xxxx
  • 单聊也能享受 Team 编排能力

3C.3 收集 Team 参数

  • front_bot:Lead Agent 名称(必须在 agent_roster 中)
  • team.roster:Specialist Agent 名称列表
  • public_updates
    • minimal — 只发 Lead 显式回复(安静,推荐)
    • normal — 加上关键事件(blocked/failed/done)
    • verbose — 所有里程碑(调试用)
  • max_parallel:最大并行任务数(默认 3)
  • auto_promote(可选):是否开启关键词自动升级到 Team 模式

3C.4 生成 Team 配置段

在 Multi-agent 配置基础上追加:

toml
1<Group 模式> 2[[group]] 3scope = "<group scope>" 4name = "<group name>" 5 6[group.mode] 7interaction = "team" 8front_bot = "<lead agent 名>" 9channel = "<lark|dingtalk|ws>" 10auto_promote = <true|false> 11 12[group.team] 13roster = ["<specialist1>", "<specialist2>"] 14public_updates = "<minimal|normal|verbose>" 15max_parallel = <N>
toml
1<Team Scope 模式(单聊)> 2[[team_scope]] 3scope = "<精确 scope>" 4name = "<name>" 5 6[team_scope.mode] 7interaction = "team" 8front_bot = "<lead agent 名>" 9channel = "<channel>" 10 11[team_scope.team] 12roster = ["<specialist1>"] 13public_updates = "minimal" 14max_parallel = 2

Phase 4:Channel 配置(可选)

询问是否需要接入 IM Channel:

是否接入消息 Channel?(Gateway 默认只有 WebSocket,可以接入 IM)

1. 飞书(Lark/Feishu)      → 运行 /add-lark
2. 钉钉(DingTalk)          → 运行 /add-dingtalk
3. 暂时不接,只用 WebSocket   → 跳过

如果用户选择接入 Channel,直接调用对应 skill:

  • 选 1:提示用户"请运行 /add-lark 继续"
  • 选 2:提示用户"请运行 /add-dingtalk 继续"

Phase 5:写入配置文件

将上述所有配置合并,写入 ~/.clawbro/config.toml

bash
1# 备份现有配置(如果存在) 2[ -f ~/.clawbro/config.toml ] && cp ~/.clawbro/config.toml ~/.clawbro/config.toml.bak.$(date +%Y%m%d%H%M%S) && echo "✓ 旧配置已备份" 3 4# 写入新配置 5cat > ~/.clawbro/config.toml << 'TOMLEOF' 6<根据 Phase 3 生成的完整 TOML 内容> 7TOMLEOF 8echo "✓ 配置已写入 ~/.clawbro/config.toml"

Phase 6:验证

6.1 先加载 env(当前 shell session)

bash
1source ~/.clawbro/.env 2echo "✓ 环境变量已加载"

6.2 语法校验

启动 gateway 做拓扑校验(不实际监听,遇错即退):

bash
1clawbro-gateway --validate-only 2>&1 | head -20

注意:如果 --validate-only 不支持,可以先启动 gateway 后立刻看启动日志。

6.3 启动 gateway 并验证

bash
1clawbro-gateway & 2GATEWAY_PID=$! 3sleep 2 4 5# 读取端口 6PORT=$(cat ~/.clawbro/gateway.port 2>/dev/null || echo "8080") 7echo "Gateway 监听在 :$PORT" 8 9# 健康检查 10curl -s http://127.0.0.1:$PORT/health | python3 -m json.tool 2>/dev/null || curl -s http://127.0.0.1:$PORT/health 11echo "" 12curl -s http://127.0.0.1:$PORT/doctor | python3 -m json.tool 2>/dev/null || curl -s http://127.0.0.1:$PORT/doctor 13echo "" 14 15# 显示 Backend 状态 16curl -s http://127.0.0.1:$PORT/diagnostics/backends 2>/dev/null 17 18kill $GATEWAY_PID 2>/dev/null

6.4 快速功能测试(可选)

如果健康检查通过,可以发送一条测试消息验证 AI 是否工作:

bash
1source ~/.clawbro/.env 2clawbro-gateway & 3GATEWAY_PID=$! 4sleep 2 5 6PORT=$(cat ~/.clawbro/gateway.port 2>/dev/null || echo "8080") 7TOKEN=$(grep 'ws_token' ~/.clawbro/config.toml 2>/dev/null | sed 's/.*= *"//' | sed 's/".*//') 8 9# 如有 token,构造 auth header 10AUTH_HEADER="" 11[ -n "$TOKEN" ] && AUTH_HEADER="-H \"Authorization: Bearer $TOKEN\"" 12 13echo "正在通过 WebSocket 发送测试消息..." 14echo '{"id":"test-1","session_key":{"channel":"ws","scope":"user:setup-test"},"content":{"type":"Text","text":"你好,请用一句话回复我"},"sender":"setup-test","channel":"ws","timestamp":"2026-01-01T00:00:00Z","thread_ts":null,"target_agent":null,"source":"human"}' | \ 15 websocat $AUTH_HEADER ws://127.0.0.1:$PORT/ws 2>/dev/null | head -5 || \ 16 echo "(websocat 未安装,跳过 WS 测试)" 17 18kill $GATEWAY_PID 2>/dev/null

Phase 7:完成与后续步骤

7.1 打印配置摘要

╔════════════════════════════════════════╗
║     ClawBro Gateway 配置完成           ║
╠════════════════════════════════════════╣
║ 配置文件: ~/.clawbro/config.toml        ║
║ 运行模式: <solo|multi-agent|team>       ║
║ 默认 Backend: clawbro-rust-agent        ║
║ Provider: <anthropic|openai|deepseek>  ║
║ Channel: <none|lark|dingtalk>          ║
╚════════════════════════════════════════╝

7.2 启动命令

bash
1# 每次启动前加载 env 2source ~/.clawbro/.env && clawbro-gateway

7.3 后续扩展

可运行的其他 Skill:
  /add-lark           — 接入飞书
  /add-dingtalk       — 接入钉钉
  /add-acp-backend    — 添加 claude-code / codex / qwen 等外部 ACP Agent
  /add-agent          — 向 roster 追加新 Agent
  /add-team-mode      — 为群组开启 Team 模式
  /doctor             — 诊断配置问题

回滚

如果配置有误,恢复备份:

bash
1BACKUP=$(ls -t ~/.clawbro/config.toml.bak.* 2>/dev/null | head -1) 2[ -n "$BACKUP" ] && cp "$BACKUP" ~/.clawbro/config.toml && echo "✓ 已恢复 $BACKUP" || echo "无备份可恢复"

Habilidades relacionadas

Looking for an alternative to setup or another community skill for your workflow? Explore these related open-source skills.

Ver todo

openclaw-release-maintainer

Logo of openclaw
openclaw

Your own personal AI assistant. Any OS. Any Platform. The lobster way. 🦞

333.8k
0
Inteligencia Artificial

widget-generator

Logo of f
f

Generate customizable widget plugins for the prompts.chat feed system

149.6k
0
Inteligencia Artificial

flags

Logo of vercel
vercel

The React Framework

138.4k
0
Navegador

pr-review

Logo of pytorch
pytorch

Tensors and Dynamic neural networks in Python with strong GPU acceleration

98.6k
0
Desarrollador