bidewio/better-openclaw

最近记录的来源活动: 2026年3月22日 17:30

SkillsMP 收录数据更新: 2026年11月12日 12:52

已收集 skills: 192

GitHub 星标: 58
GitHub Forks: 6

用 AI 助手安装

复制这段 Prompt，发给你正在使用的 AI 助手。

Follow https://skillsmp.com/skill-install/prompt.md to install Agent Skills from https://github.com/bidewio/better-openclaw.

仓库 README

仓库 README 来源：README.md

查看 README 源文件

展开完整 README

<h1 align="center">better-openclaw</h1> Build your AI agent superstack in seconds. 94 services. 8 agent frameworks. 10 skill packs. 9 presets. One command. <a href="#quick-start">Quick Start</a> • <a href="#features">Features</a> • <a href="#deployment">Deployment</a> • <a href="#service-catalog">Services</a> • <a href="#skill-packs">Skill Packs</a> • <a href="#presets">Presets</a> • <a href="#rest-api">API</a> • <a href="#architecture">Architecture</a> • <a href="#development">Development</a> <img alt="License" src="https://img.shields.io/badge/license-AGPL--3.0-blue.svg" /> <img alt="Node" src="https://img.shields.io/badge/node-%3E%3D22-brightgreen.svg" /> <img alt="pnpm" src="https://img.shields.io/badge/pnpm-10.30.3-orange.svg" /> <img alt="TypeScript" src="https://img.shields.io/badge/TypeScript-5.7-blue.svg" /> --- **better-openclaw** is a CLI tool, REST API, and web UI for scaffolding production-ready AI agent stacks with Docker Compose. Choose from **8 agent frameworks** (OpenClaw, CoPaw, NanoClaw, NanoBot, ZeroClaw, MemU, Claude Code, Codex), pick your services, choose skill packs, and get a fully wired `docker-compose.yml`, `.env`, reverse proxy configs, monitoring dashboards, and agent skill files -- all in one command. ## Quick Start ```bash npx create-better-openclaw@latest ``` Follow the interactive wizard to select services, skill packs, and configuration options. Your complete stack will be generated in seconds. Or run non-interactively: ```bash # Use a preset npx create-better-openclaw --preset researcher --yes # Cherry-pick services npx create-better-openclaw --services postgresql,redis,n8n,grafana --proxy caddy --domain example.com --yes # Use a different agent framework npx create-better-openclaw --preset researcher --primary-framework zeroclaw --yes # Run multiple frameworks side-by-side npx create-better-openclaw --services postgresql,redis --primary-framework memu --companion-frameworks copaw,nanoclaw --yes # Preview without writing files npx create-better-openclaw --preset devops --dry-run # JSON output for CI/CD pipelines npx create-better-openclaw --preset minimal --yes --json ``` ### CLI Commands ```bash npx create-better-openclaw generate [dir] # Generate a stack (default command) npx create-better-openclaw services list # List all 94 available services npx create-better-openclaw presets list # List all preset stacks npx create-better-openclaw presets info <id> # Show preset details npx create-better-openclaw validate <dir> # Validate an existing stack npx create-better-openclaw init # Initialize in current directory npx create-better-openclaw add <service-id> # Add a service to existing stack npx create-better-openclaw remove <service-id> # Remove a service from existing stack npx create-better-openclaw deploy # Deploy stack to Dokploy or Coolify (interactive) npx create-better-openclaw deploy --provider dokploy --url https://... --api-key ... # Non-interactive npx create-better-openclaw completion <shell> # Generate shell completions (bash/zsh/fish) ``` ### Advanced CLI Options ```bash # Custom proxy ports (auto-detects conflicts) npx create-better-openclaw --proxy caddy --proxy-http-port 8080 --proxy-https-port 8443 # Select AI providers (comma-separated) npx create-better-openclaw --ai-providers openai,anthropic,google --yes # Combine options npx create-better-openclaw \ --services postgresql,redis,n8n \ --proxy caddy --proxy-http-port 8080 \ --ai-providers openai,anthropic \ --domain example.com \ --yes ``` ## Features - **8 agent frameworks** -- choose OpenClaw, CoPaw, NanoClaw, NanoBot, ZeroClaw, MemU, Claude Code, or Codex as your primary orchestrator, with optional companion frameworks for hybrid stacks - **Interactive CLI wizard** -- guided service selection with dependency resolution - **Non-interactive mode** -- scriptable with presets and flags for CI/CD pipelines - **Automatic port conflict detection** -- scans your system for port conflicts and auto-reassigns services to available ports (interactive and non-interactive modes) - **REST API** -- generate stacks programmatically via HTTP endpoints - **Web UI** -- visual stack builder with live preview and one-click download - **Smart dependency resolution** -- automatically adds required services (e.g., n8n adds PostgreSQL) - **Preset stacks** -- pre-configured templates for common use cases - **Skill packs** -- bundles of agent skills wired to their backing services - **Reverse proxy configs** -- auto-generated Caddy or Traefik configuration with label generation and custom port support - **Auto-generated OpenAPI spec** -- live Swagger UI at `/api/v1/docs` - **Distributed rate limiting** -- optional Redis-backed rate limiter for production - **Config migrations** -- forward-compatible configuration versioning - **Monitoring dashboards** -- Grafana + Prometheus pre-wired with service exporters - **Environment files** -- secure `.env` generation with random secrets - **One-click deploy** -- deploy generated stacks to self-hosted Dokploy or Coolify directly from the web UI or CLI - **Centralized operations logging** -- structured NDJSON logs across all packages (CLI, API, MCP) with correlation IDs, step tracking, duration measurement, and automatic sensitive data redaction - **Bridge sidecar** -- REST API for live stack management (service status, logs, restart/stop/scale commands) with Convex heartbeat sync to Mission Control - **Validation engine** -- port conflicts, resource limits, and configuration checks - **Multi-platform support** -- linux/amd64 and linux/arm64 - **Bare-metal deployment** -- hybrid native + Docker: run supported services natively on the host and use Docker only for the rest (see [Deployment](#deployment)) ## Deployment You can generate stacks for **Docker** (default) or **bare-metal**: - **Docker** — All services run in containers. Use `docker compose up` to start everything. - **Bare-metal** — A **native + Docker hybrid**: services that have a **native recipe** (install/run on the host) run natively; the rest run in Docker. The generator outputs: - One **docker-compose** for Docker-only services plus the OpenClaw gateway. - **Native install/run scripts** (e.g. `native/install-linux.sh`) for services that support native install on the chosen platform. - A **top-level installer** (`install.sh` or `install.ps1`) that installs and starts native services first, then runs `docker compose up` for the rest. Only services with a native recipe run on the host; others remain in Docker. Currently **Redis** supports a native Linux recipe (apt/dnf + systemd). More services (e.g. PostgreSQL, Caddy, Prometheus) may be added over time. Node/Python apps, La Suite Meet, Ollama, and similar stay Docker-only. ## Service Catalog 94 services across 21 categories, ready to compose: | Category | Services | |---|---| | **AI Coding Agents** | Claude Code, Codex, Gemini CLI, Kimi, OpenCode | | **AI Platforms & Chat UIs** | AnythingLLM, Dify, Flowise, LibreChat, LiteLLM, Open WebUI | | **AI / Local Models** | ComfyUI, Ollama, Stable Diffusion, Whisper | | **Automation & Workflows** | Cal.com, Home Assistant, n8n, Temporal, xyOps | | **Database & Caching** | Convex, Neo4j, PostgreSQL, Redis, Supabase, Valkey | | **Vector Databases** | ChromaDB, Milvus, Qdrant, Weaviate | | **Media & Video** | FFmpeg, Immich, Jellyfin, Motion Canvas, Remotion | | **Social Media** | Ghost, Mixpost, Postiz | | **Analytics** | Matomo, OpenPanel, Umami | | **Knowledge & Documents** | AppFlowy, DocsGPT, NocoDB, Outline, Paperless-ngx | | **Object Storage** | MinIO, Nextcloud | | **Developer Tools** | Beszel, Bridge, Code Server, Convex Dashboard, Coolify, Dokploy, Dozzle, Gitea, Headscale, Jenkins, Mission Control, Portainer, Tailscale, Watchtower | | **Reverse Proxy** | Caddy, Traefik | | **Monitoring** | Grafana, Loki, Prometheus, SigNoz, Uptime Kuma | | **Browser Automation** | Browserless, LightPanda, Playwright Server, Scrapling, Steel Browser | | **Search** | Meilisearch, SearXNG | | **Communication** | Gotify, La Suite Meet (frontend/backend/agents), LiveKit, Matrix Synapse, Mattermost, ntfy, Rocket.Chat, UseSend | | **Streaming** | Stream Gateway | | **Security** | Authentik, CrowdSec, HexStrike, PentAGI, PentestAgent, SolidityGuard, Vaultwarden | | **Desktop** | Desktop Environment (KasmVNC) | Every service definition includes a pinned Docker image tag, ports, volumes, health checks, environment variables, resource limits, and dependency declarations. ## Skill Packs Skill packs bundle agent skills with their required infrastructure: | Pack | Description | Services | |---|---|---| | **Video Creator** | Create and process videos programmatically | FFmpeg, Remotion, MinIO | | **Research Agent** | Web research with vector memory and scraping | Qdrant, SearXNG, Browserless | | **Social Media** | Content processing with caching and storage | FFmpeg, Redis, MinIO | | **DevOps** | Monitoring, automation, and alerting | n8n, Redis, Uptime Kuma, Grafana, Prometheus | | **Knowledge Base** | Document indexing with vector + full-text search | Qdrant, PostgreSQL, Meilisearch | | **Local AI** | Local LLM inference and speech-to-text | Ollama, Whisper | | **Content Creator** | AI-powered image and video generation | FFmpeg, Remotion, MinIO, Stable Diffusion | | **AI Playground** | Multi-model AI experimentation | Ollama, Open WebUI, Qdrant, LiteLLM | | **Coding Team** | AI coding agents with shared state | Claude Code, Codex, Redis, PostgreSQL | | **Knowledge Hub** | Wiki and document search | Outline, Qdrant, Meilisearch, PostgreSQL | ## Presets Pre-configured stack templates for quick starts: | Preset | Services | Memory | |---|---|---| | **Minimal** | Redis | ~1 GB | | **Creator** | FFmpeg, Remotion, MinIO, Redis | ~2 GB | | **Researcher** | Qdrant, SearXNG, Browserless, Redis | ~2.5 GB | | **DevOps** | n8n, PostgreSQL, Redis, Uptime Kuma, Grafana, Prometheus | ~3 GB | | **Content Creator** | FFmpeg, Remotion, MinIO, Redis, Stable Diffusion | ~4 GB | | **AI Playground** | Ollama, Open WebUI, Qdrant, LiteLLM, Redis | ~6 GB | | **Coding Team** | Claude Code, Codex, Redis, PostgreSQL | ~3 GB | | **La Suite Meet** | Meet frontend/backend/agents, Redis, PostgreSQL, LiveKit | ~4 GB | | **Full Stack** | All core services + all skill packs | ~8 GB | ## Agent Frameworks Choose your primary agent orchestrator and optionally add companion frameworks for hybrid stacks: | Framework | ID | Description | |---|---|---| | **OpenClaw** | `openclaw` | Full-featured agent framework with gateway, mission control, and Tailscale networking (default) | | **CoPaw** | `copaw` | Cooperative multi-agent framework with shared memory | | **NanoClaw** | `nanoclaw` | Lightweight single-agent runtime for resource-constrained environments | | **NanoBot** | `nanobot` | Minimal bot framework with plugin architecture | | **ZeroClaw** | `zeroclaw` | Zero-config agent framework with convention-over-configuration | | **MemU** | `memu` | Memory-first agent framework with persistent context and PostgreSQL backing | | **Claude Code** | `claude-code` | Anthropic's CLI agent for software engineering tasks | | **Codex** | `codex` | OpenAI's CLI agent for code generation and editing | Non-OpenClaw frameworks skip Convex, Mission Control, and Tailscale. Each framework generates its own gateway container, CLI services, and network configuration. ## REST API The API runs on port 3456 with auto-generated OpenAPI docs at `/api/v1/docs`. ```bash # List all services curl http://localhost:3456/api/v1/services # Get a specific service curl http://localhost:3456/api/v1/services/postgresql # Filter services by category curl http://localhost:3456/api/v1/services?category=database # List presets curl http://localhost:3456/api/v1/presets # Get preset with resolved service details curl http://localhost:3456/api/v1/presets/devops # Generate a stack curl -X POST http://localhost:3456/api/v1/generate \ -H "Content-Type: application/json" \ -d '{"services": ["postgresql", "redis", "n8n"], "proxy": "caddy"}' # Generate with a specific agent framework curl -X POST http://localhost:3456/api/v1/generate \ -H "Content-Type: application/json" \ -d '{"services": ["postgresql", "redis"], "primaryFramework": "zeroclaw", "companionFrameworks": ["copaw"]}' ``` API key authentication is supported via the `X-API-Key` header. Configure allowed keys with the `API_KEYS` environment variable (comma-separated). ## Architecture Main workspace packages and repo tooling: ``` better-openclaw/ ├── packages/ │ ├── core/ # Schemas, service registry, resolver, composer, validators │ ├── cli/ # Interactive wizard + non-interactive CLI (Commander + Clack) │ ├── api/ # REST API (Hono + Zod OpenAPI) │ ├── web/ # Web UI (Next.js 16 + Tailwind CSS 4) │ ├── db/ # Shared Drizzle schema and database client │ ├── mcp/ # MCP server for agent integrations │ └── mission-control/ # Vite + Convex operations dashboard ├── turbo.json # Turborepo task pipeline ├── biome.json # Linting and formatting (Biome) └── vitest.config.ts # Shared Vitest defaults for non-UI packages ``` ### Package Dependency Graph - `core` powers the CLI, API, web app, and MCP server. - `db` is shared by the API/auth layer for persisted user data. - `mission-control` is a separate Vite + Convex app within the same workspace. ### Key Technologies - **Runtime**: Node.js >= 22 - **Language**: TypeScript 5.7 - **Monorepo**: pnpm workspaces + Turborepo - **API**: Hono with Zod OpenAPI - **Web**: Next.js 16 with React 19, Tailwind CSS 4, Framer Motion - **CLI**: Commander + @clack/prompts - **Validation**: Zod schemas throughout - **Testing**: Vitest - **Linting**: Biome - **Build**: tsdown (ESM + CJS dual-format output) ## Development ### Prerequisites - Node.js >= 22 - pnpm 10.30.3 ### Setup ```bash # Clone the repository git clone https://github.com/bidewio/better-openclaw.git cd better-openclaw # Install dependencies pnpm install # Start the main local apps in development mode pnpm dev ``` This starts: - **API** at `http://localhost:3456` (with hot reload via tsx) - **Web** at `http://localhost:3654` (with Next.js fast refresh) - **Mission Control** at `http://localhost:3660` (with Vite + Convex dev tooling) Run `pnpm dev:mcp` to start the MCP server separately. `packages/db` is a shared library package and exposes `db:*` scripts for schema and migration work rather than a long-running dev server. ### Commands | Command | Description | |---|---| | `pnpm dev` | Start the main local apps in dev mode | | `pnpm dev:mcp` | Start the MCP server in dev mode | | `pnpm dev:all` | Start the main apps plus the MCP server | | `pnpm build` | Build all packages | | `pnpm test` | Run all tests | | `pnpm lint` | Lint all packages | | `pnpm typecheck` | Type-check all packages | | `pnpm format` | Format all files with Biome | | `pnpm clean` | Remove all build artifacts | ### Docker Build and run the full stack with Docker: ```bash docker compose up --build ``` This exposes: - **Web UI** at `http://localhost:3000` - **API** at `http://localhost:3456` ## Contributing Contributions are welcome! Please see the [Contributing Guide](CONTRIBUTING.md) for details on: - Adding new service definitions - Creating skill packs - Building preset stacks - Writing tests - Code style and conventions ## License [AGPL-3.0](LICENSE) --- Made with care by <a href="https://bidew.io">bachir@bidew.io</a>

正在加载仓库 README…

这个仓库中的 skills

软件开发工程师网络与计算机系统管理员网页开发工程师信息安全分析师

14 个职业分类 · 已分类 100%

网络与计算机系统管理员

Configuration management and automation with Ansible

原文语言：英语

2026-02-21

已展示 40 / 192 个已收集 Skill。