[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"project-74222":3},{"id":4,"name":5,"fullName":6,"owner":7,"repo":5,"description":8,"homepage":9,"htmlUrl":10,"language":11,"languages":10,"totalLinesOfCode":10,"stars":12,"forks":13,"watchers":14,"openIssues":15,"contributorsCount":16,"subscribersCount":16,"size":16,"stars1d":17,"stars7d":18,"stars30d":19,"stars90d":16,"forks30d":16,"starsTrendScore":20,"compositeScore":21,"rankGlobal":10,"rankLanguage":10,"license":22,"archived":23,"fork":23,"defaultBranch":24,"hasWiki":25,"hasPages":25,"topics":26,"createdAt":10,"pushedAt":10,"updatedAt":32,"readmeContent":33,"aiSummary":34,"trendingCount":16,"starSnapshotCount":16,"syncStatus":15,"lastSyncTime":35,"discoverSource":36},74222,"build-your-own-openclaw","czl9707\u002Fbuild-your-own-openclaw","czl9707","A step-by-step guide to build your own AI agent.","https:\u002F\u002Fbuild-your-own-openclaw.kiyo-n-zane.com\u002F",null,"Python",1708,305,13,2,0,11,26,99,33,98.36,"MIT License",false,"main",true,[27,28,29,30,31],"ai-agent","build-your-own-x","llm","python","tutorial","2026-06-12 04:01:13","# Build Your Own OpenClaw\n\nA step-by-step tutorial to build your own AI agent, from a simple chat loop to a lightweight version of [OpenClaw](https:\u002F\u002Fgithub.com\u002Fopenclaw\u002Fopenclaw).\n\n\u003Cimg src=\"Cover.png\" style=\"width: 100%;\">\n\n## Overview\n\n**18 progressive steps** that teach you how to build an minimal version of OpenClaw. Each step includes:\n\n- A `README.md` going through key components and design decision.\n- A Runnable codebase.\n\n**Example Project:** [pickle-bot](https:\u002F\u002Fgithub.com\u002Fczl9707\u002Fpickle-bot) - our reference implementation\n\n## Tutorial Structure\n\n### Phase 1: Capable Single Agent (Steps 0-6)\nBuild a fully-functional agent that can chat, use tools, learn skills, remember conversations, and access the internet.\n\n- [**00-chat-loop**](.\u002F00-chat-loop\u002F) - Just a Chat Loop\n- [**01-tools**](.\u002F01-tools\u002F) - Give your agent a tool.\n- [**02-skills**](.\u002F02-skills\u002F) - Extend your agent with `SKILL.md`\n- [**03-persistence**](.\u002F03-persistence\u002F) - Save your conversations.\n- [**04-slash-commands**](.\u002F04-slash-commands\u002F) - Direct user control over sessions.\n- [**05-compaction**](.\u002F05-compaction\u002F) - Pack you history and carry on...\n- [**06-web-tools**](.\u002F06-web-tools\u002F) - Your Agent want to see the bigger world.\n\n### Phase 2: Event-Driven Architecture (Steps 7-10)\nRefactor to event-driven architecture for scalability and multi-platform support.\n\n- [**07-event-driven**](.\u002F07-event-driven\u002F) - Expose you agent beyond CLI.\n- [**08-config-hot-reload**](.\u002F08-config-hot-reload\u002F) - Edit without restart.\n- [**09-channels**](.\u002F09-channels\u002F) - Talk to your agent from on your phone.\n- [**10-websocket**](.\u002F10-websocket\u002F) - Want to interact with you agent programatically?\n\n### Phase 3: Autonomous & Multi-Agent (Steps 11-15)\nAdd scheduled tasks, agent collaboration, and intelligent routing.\n\n- [**11-multi-agent-routing**](.\u002F11-multi-agent-routing\u002F) - Route right job to right agent.\n- [**12-cron-heartbeat**](.\u002F12-cron-heartbeat\u002F) - An Agent work while you are sleeping.\n- [**13-multi-layer-prompts**](.\u002F13-multi-layer-prompts\u002F) - More Context, More Context, More Context.\n- [**14-post-message-back**](.\u002F14-post-message-back\u002F) - Your Agent want to Speak to you.\n- [**15-agent-dispatch**](.\u002F15-agent-dispatch\u002F) - Your Agent want friends to work with!\n\n### Phase 4: Production & Scale (Steps 16-17)\nFeatures for reliability and long-term memory.\n\n- [**16-concurrency-control**](.\u002F16-concurrency-control\u002F) - Too many Pickle are running at the same time?\n- [**17-memory**](.\u002F17-memory\u002F) - Remember me!\n\n## How to Use This Tutorial\n\n### Configure API Keys\n\nBefore running any step, you need to configure your API keys:\n\n1. **Copy the example config:**\n   ```bash\n   cp default_workspace\u002Fconfig.example.yaml default_workspace\u002Fconfig.user.yaml\n   ```\n\n2. **Edit `config.user.yaml`** with your API keys:\n   - See [LiteLLM providers](https:\u002F\u002Fdocs.litellm.ai\u002Fdocs\u002Fproviders) for the full list of supported providers\n   - Check out [Provider Examples](PROVIDER_EXAMPLES.md) for some examples\n\n3. Just follow each steps, read and try it out.\n\n## Contributing\n\nEach step is implemented in a separate session. Feel free to suggest improvements!\n","该项目提供了一个逐步指南，帮助用户从零开始构建自己的AI代理。通过18个渐进步骤，涵盖了从简单的聊天循环到轻量级OpenClaw版本的开发过程。每个步骤都包括详细的文档说明和可运行代码示例，采用Python语言编写。项目强调了如何为代理添加工具、技能、持久化存储以及事件驱动架构等核心功能，并支持多平台扩展与部署。适用于希望深入了解AI代理内部机制并亲手实践构建的开发者或研究者，在教学和个人项目中尤为适用。","2026-06-11 03:49:34","high_star"]