[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M7C3396KGPDVP9JJK40AXQJA":3,"skill-same:01M7C3396KGPDVP9JJK40AXQJA":44},{"skillNo":4,"name":5,"nameZh":6,"description":7,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":16,"categories":17,"keywords":18,"publishedAt":19,"linkOnly":20,"markdownContent":21,"markdownBody":22,"frontMatter":23,"refName":24,"commitSha":25,"skillDirectory":26,"forksCount":14,"repositoryHtmlUrl":27,"repositoryHomepageUrl":6,"ownerHtmlUrl":28,"ownerBlogUrl":6,"ownerDescription":6,"updatedAt":29,"install":30,"resources":43},"01M7C3396KGPDVP9JJK40AXQJA","google-adk-python",null,"Build AI agents with Google ADK Python (Agent Development Kit). Use for multi-agent systems, workflow agents (sequential\u002Fparallel\u002Floop), Vertex AI deployment, tool integration, human-in-the-loop.","Truc350\u002FWebHomeStay","Truc350","https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F176447042?v=4","01M7C2S7D6AAK15J66E3NDXSXK","Huỳnh Thị Thanh Trúc","01M7C2S7DTPMN5ZM8876HF8KF8",0,"MIT","https:\u002F\u002Fgithub.com\u002FTruc350\u002FWebHomeStay\u002Fblob\u002Fmain\u002F.agents\u002Fskills\u002Fgoogle-adk-python\u002FSKILL.md",[],[],"2026-11-15 14:19:17",false,"---\nname: google-adk-python\ndescription: Build AI agents with Google ADK Python (Agent Development Kit). Use for multi-agent systems, workflow agents (sequential\u002Fparallel\u002Floop), Vertex AI deployment, tool integration, human-in-the-loop.\nlicense: MIT\n---\n\n# Google ADK Python Skill\n\nYou are an expert guide for Google's Agent Development Kit (ADK) Python - an open-source, code-first toolkit for building, evaluating, and deploying AI agents.\n\n## When to Use This Skill\n\nUse this skill when users need to:\n- Build AI agents with tool integration and orchestration capabilities\n- Create multi-agent systems with hierarchical coordination\n- Implement workflow agents (sequential, parallel, loop) for predictable pipelines\n- Integrate LLM-powered agents with Google Search, Code Execution, or custom tools\n- Deploy agents to Vertex AI Agent Engine, Cloud Run, or custom infrastructure\n- Evaluate and test agent performance systematically\n- Implement human-in-the-loop approval flows for tool execution\n\n## Core Concepts\n\n### Agent Types\n\n**LlmAgent**: LLM-powered agents capable of dynamic routing and adaptive behavior\n- Define with name, model, instruction, description, and tools\n- Supports sub-agents for delegation and coordination\n- Intelligent decision-making based on context\n\n**Workflow Agents**: Structured, predictable orchestration patterns\n- **SequentialAgent**: Execute agents in defined order\n- **ParallelAgent**: Run multiple agents concurrently\n- **LoopAgent**: Repeat execution with iteration logic\n\n**BaseAgent**: Foundation for custom agent implementations\n\n### Key Components\n\n**Tools Ecosystem**:\n- Pre-built tools (google_search, code_execution)\n- Custom Python functions as tools\n- OpenAPI specification integration\n- Tool confirmation flows for human approval\n\n**Multi-Agent Architecture**:\n- Hierarchical agent composition\n- Specialized agents for specific domains\n- Coordinator agents for delegation\n\n## Installation\n\n```bash\n# Stable release (recommended)\npip install google-adk\n\n# Development version (latest features)\npip install git+https:\u002F\u002Fgithub.com\u002Fgoogle\u002Fadk-python.git@main\n```\n\n## Implementation Patterns\n\n### Single Agent with Tools\n\n```python\nfrom google.adk.agents import LlmAgent\nfrom google.adk.tools import google_search\n\nagent = LlmAgent(\n    name=\"search_assistant\",\n    model=\"gemini-2.5-flash\",\n    instruction=\"You are a helpful assistant that searches the web for information.\",\n    description=\"Search assistant for web queries\",\n    tools=[google_search]\n)\n```\n\n### Multi-Agent System\n\n```python\nfrom google.adk.agents import LlmAgent\n\n# Specialized agents\nresearcher = LlmAgent(\n    name=\"Researcher\",\n    model=\"gemini-2.5-flash\",\n    instruction=\"Research topics thoroughly using web search.\",\n    tools=[google_search]\n)\n\nwriter = LlmAgent(\n    name=\"Writer\",\n    model=\"gemini-2.5-flash\",\n    instruction=\"Write clear, engaging content based on research.\",\n)\n\n# Coordinator agent\ncoordinator = LlmAgent(\n    name=\"Coordinator\",\n    model=\"gemini-2.5-flash\",\n    instruction=\"Delegate tasks to researcher and writer agents.\",\n    sub_agents=[researcher, writer]\n)\n```\n\n### Custom Tool Creation\n\n```python\nfrom google.adk.tools import Tool\n\ndef calculate_sum(a: int, b: int) -> int:\n    \"\"\"Calculate the sum of two numbers.\"\"\"\n    return a + b\n\n# Convert function to tool\nsum_tool = Tool.from_function(calculate_sum)\n\nagent = LlmAgent(\n    name=\"calculator\",\n    model=\"gemini-2.5-flash\",\n    tools=[sum_tool]\n)\n```\n\n### Sequential Workflow\n\n```python\nfrom google.adk.agents import SequentialAgent\n\nworkflow = SequentialAgent(\n    name=\"research_workflow\",\n    agents=[researcher, summarizer, writer]\n)\n```\n\n### Parallel Workflow\n\n```python\nfrom google.adk.agents import ParallelAgent\n\nparallel_research = ParallelAgent(\n    name=\"parallel_research\",\n    agents=[web_researcher, paper_researcher, expert_researcher]\n)\n```\n\n### Human-in-the-Loop\n\n```python\nfrom google.adk.tools import google_search\n\n# Tool with confirmation required\nagent = LlmAgent(\n    name=\"careful_searcher\",\n    model=\"gemini-2.5-flash\",\n    tools=[google_search],\n    tool_confirmation=True  # Requires approval before execution\n)\n```\n\n## Deployment Options\n\n### Cloud Run Deployment\n\n```bash\n# Containerize agent\ndocker build -t my-agent .\n\n# Deploy to Cloud Run\ngcloud run deploy my-agent --image my-agent\n```\n\n### Vertex AI Agent Engine\n\n```python\n# Deploy to Vertex AI for scalable agent hosting\n# Integrates with Google Cloud's managed infrastructure\n```\n\n### Custom Infrastructure\n\n```python\n# Run agents locally or on custom servers\n# Full control over deployment environment\n```\n\n## Model Support\n\n**Optimized for Gemini**:\n- gemini-2.5-flash\n- gemini-2.5-pro\n- gemini-1.5-flash\n- gemini-1.5-pro\n\n**Model Agnostic**: While optimized for Gemini, ADK supports other LLM providers through standard APIs.\n\n## Best Practices\n\n1. **Code-First Philosophy**: Define agents in Python for version control, testing, and flexibility\n2. **Modular Design**: Create specialized agents for specific domains, compose into systems\n3. **Tool Integration**: Leverage pre-built tools, extend with custom functions\n4. **Evaluation**: Test agents systematically against test cases\n5. **Safety**: Implement confirmation flows for sensitive operations\n6. **Hierarchical Structure**: Use coordinator agents for complex multi-agent workflows\n7. **Workflow Selection**: Choose workflow agents for predictable pipelines, LLM agents for dynamic routing\n\n## Common Use Cases\n\n- **Research Assistants**: Web search + summarization + report generation\n- **Code Assistants**: Code execution + documentation + debugging\n- **Customer Support**: Query routing + knowledge base + escalation\n- **Content Creation**: Research + writing + editing pipelines\n- **Data Analysis**: Data fetching + processing + visualization\n- **Task Automation**: Multi-step workflows with conditional logic\n\n## Development UI\n\nADK includes built-in interface for:\n- Testing agent behavior interactively\n- Debugging tool calls and responses\n- Evaluating agent performance\n- Iterating on agent design\n\n## Resources\n\n- GitHub: https:\u002F\u002Fgithub.com\u002Fgoogle\u002Fadk-python\n- Documentation: https:\u002F\u002Fgoogle.github.io\u002Fadk-docs\u002F\n- llms.txt: https:\u002F\u002Fraw.githubusercontent.com\u002Fgoogle\u002Fadk-python\u002Frefs\u002Fheads\u002Fmain\u002Fllms.txt\n\n## Implementation Workflow\n\nWhen implementing ADK-based agents:\n\n1. **Define Requirements**: Identify agent capabilities and tools needed\n2. **Choose Architecture**: Single agent, multi-agent, or workflow-based\n3. **Select Tools**: Pre-built, custom functions, or OpenAPI integrations\n4. **Implement Agents**: Create agent definitions with instructions and tools\n5. **Test Locally**: Use development UI for iteration\n6. **Add Evaluation**: Create test cases for systematic validation\n7. **Deploy**: Choose Cloud Run, Vertex AI, or custom infrastructure\n8. **Monitor**: Track agent performance and iterate\n\nRemember: ADK treats agent development like traditional software engineering - use version control, write tests, and follow engineering best practices.","# Google ADK Python Skill\n\nYou are an expert guide for Google's Agent Development Kit (ADK) Python - an open-source, code-first toolkit for building, evaluating, and deploying AI agents.\n\n## When to Use This Skill\n\nUse this skill when users need to:\n- Build AI agents with tool integration and orchestration capabilities\n- Create multi-agent systems with hierarchical coordination\n- Implement workflow agents (sequential, parallel, loop) for predictable pipelines\n- Integrate LLM-powered agents with Google Search, Code Execution, or custom tools\n- Deploy agents to Vertex AI Agent Engine, Cloud Run, or custom infrastructure\n- Evaluate and test agent performance systematically\n- Implement human-in-the-loop approval flows for tool execution\n\n## Core Concepts\n\n### Agent Types\n\n**LlmAgent**: LLM-powered agents capable of dynamic routing and adaptive behavior\n- Define with name, model, instruction, description, and tools\n- Supports sub-agents for delegation and coordination\n- Intelligent decision-making based on context\n\n**Workflow Agents**: Structured, predictable orchestration patterns\n- **SequentialAgent**: Execute agents in defined order\n- **ParallelAgent**: Run multiple agents concurrently\n- **LoopAgent**: Repeat execution with iteration logic\n\n**BaseAgent**: Foundation for custom agent implementations\n\n### Key Components\n\n**Tools Ecosystem**:\n- Pre-built tools (google_search, code_execution)\n- Custom Python functions as tools\n- OpenAPI specification integration\n- Tool confirmation flows for human approval\n\n**Multi-Agent Architecture**:\n- Hierarchical agent composition\n- Specialized agents for specific domains\n- Coordinator agents for delegation\n\n## Installation\n\n```bash\n# Stable release (recommended)\npip install google-adk\n\n# Development version (latest features)\npip install git+https:\u002F\u002Fgithub.com\u002Fgoogle\u002Fadk-python.git@main\n```\n\n## Implementation Patterns\n\n### Single Agent with Tools\n\n```python\nfrom google.adk.agents import LlmAgent\nfrom google.adk.tools import google_search\n\nagent = LlmAgent(\n    name=\"search_assistant\",\n    model=\"gemini-2.5-flash\",\n    instruction=\"You are a helpful assistant that searches the web for information.\",\n    description=\"Search assistant for web queries\",\n    tools=[google_search]\n)\n```\n\n### Multi-Agent System\n\n```python\nfrom google.adk.agents import LlmAgent\n\n# Specialized agents\nresearcher = LlmAgent(\n    name=\"Researcher\",\n    model=\"gemini-2.5-flash\",\n    instruction=\"Research topics thoroughly using web search.\",\n    tools=[google_search]\n)\n\nwriter = LlmAgent(\n    name=\"Writer\",\n    model=\"gemini-2.5-flash\",\n    instruction=\"Write clear, engaging content based on research.\",\n)\n\n# Coordinator agent\ncoordinator = LlmAgent(\n    name=\"Coordinator\",\n    model=\"gemini-2.5-flash\",\n    instruction=\"Delegate tasks to researcher and writer agents.\",\n    sub_agents=[researcher, writer]\n)\n```\n\n### Custom Tool Creation\n\n```python\nfrom google.adk.tools import Tool\n\ndef calculate_sum(a: int, b: int) -> int:\n    \"\"\"Calculate the sum of two numbers.\"\"\"\n    return a + b\n\n# Convert function to tool\nsum_tool = Tool.from_function(calculate_sum)\n\nagent = LlmAgent(\n    name=\"calculator\",\n    model=\"gemini-2.5-flash\",\n    tools=[sum_tool]\n)\n```\n\n### Sequential Workflow\n\n```python\nfrom google.adk.agents import SequentialAgent\n\nworkflow = SequentialAgent(\n    name=\"research_workflow\",\n    agents=[researcher, summarizer, writer]\n)\n```\n\n### Parallel Workflow\n\n```python\nfrom google.adk.agents import ParallelAgent\n\nparallel_research = ParallelAgent(\n    name=\"parallel_research\",\n    agents=[web_researcher, paper_researcher, expert_researcher]\n)\n```\n\n### Human-in-the-Loop\n\n```python\nfrom google.adk.tools import google_search\n\n# Tool with confirmation required\nagent = LlmAgent(\n    name=\"careful_searcher\",\n    model=\"gemini-2.5-flash\",\n    tools=[google_search],\n    tool_confirmation=True  # Requires approval before execution\n)\n```\n\n## Deployment Options\n\n### Cloud Run Deployment\n\n```bash\n# Containerize agent\ndocker build -t my-agent .\n\n# Deploy to Cloud Run\ngcloud run deploy my-agent --image my-agent\n```\n\n### Vertex AI Agent Engine\n\n```python\n# Deploy to Vertex AI for scalable agent hosting\n# Integrates with Google Cloud's managed infrastructure\n```\n\n### Custom Infrastructure\n\n```python\n# Run agents locally or on custom servers\n# Full control over deployment environment\n```\n\n## Model Support\n\n**Optimized for Gemini**:\n- gemini-2.5-flash\n- gemini-2.5-pro\n- gemini-1.5-flash\n- gemini-1.5-pro\n\n**Model Agnostic**: While optimized for Gemini, ADK supports other LLM providers through standard APIs.\n\n## Best Practices\n\n1. **Code-First Philosophy**: Define agents in Python for version control, testing, and flexibility\n2. **Modular Design**: Create specialized agents for specific domains, compose into systems\n3. **Tool Integration**: Leverage pre-built tools, extend with custom functions\n4. **Evaluation**: Test agents systematically against test cases\n5. **Safety**: Implement confirmation flows for sensitive operations\n6. **Hierarchical Structure**: Use coordinator agents for complex multi-agent workflows\n7. **Workflow Selection**: Choose workflow agents for predictable pipelines, LLM agents for dynamic routing\n\n## Common Use Cases\n\n- **Research Assistants**: Web search + summarization + report generation\n- **Code Assistants**: Code execution + documentation + debugging\n- **Customer Support**: Query routing + knowledge base + escalation\n- **Content Creation**: Research + writing + editing pipelines\n- **Data Analysis**: Data fetching + processing + visualization\n- **Task Automation**: Multi-step workflows with conditional logic\n\n## Development UI\n\nADK includes built-in interface for:\n- Testing agent behavior interactively\n- Debugging tool calls and responses\n- Evaluating agent performance\n- Iterating on agent design\n\n## Resources\n\n- GitHub: https:\u002F\u002Fgithub.com\u002Fgoogle\u002Fadk-python\n- Documentation: https:\u002F\u002Fgoogle.github.io\u002Fadk-docs\u002F\n- llms.txt: https:\u002F\u002Fraw.githubusercontent.com\u002Fgoogle\u002Fadk-python\u002Frefs\u002Fheads\u002Fmain\u002Fllms.txt\n\n## Implementation Workflow\n\nWhen implementing ADK-based agents:\n\n1. **Define Requirements**: Identify agent capabilities and tools needed\n2. **Choose Architecture**: Single agent, multi-agent, or workflow-based\n3. **Select Tools**: Pre-built, custom functions, or OpenAPI integrations\n4. **Implement Agents**: Create agent definitions with instructions and tools\n5. **Test Locally**: Use development UI for iteration\n6. **Add Evaluation**: Create test cases for systematic validation\n7. **Deploy**: Choose Cloud Run, Vertex AI, or custom infrastructure\n8. **Monitor**: Track agent performance and iterate\n\nRemember: ADK treats agent development like traditional software engineering - use version control, write tests, and follow engineering best practices.",{"name":5,"license":15,"description":7},"main","af64bb75a16e8bc1847a568059b9300e09f27c4a",".agents\u002Fskills\u002Fgoogle-adk-python","https:\u002F\u002Fgithub.com\u002FTruc350\u002FWebHomeStay","https:\u002F\u002Fgithub.com\u002FTruc350","2026-09-04 21:59:58",{"source":8,"skillName":5,"commands":31,"prompt":41,"zipAvailable":42},[32,35,38],{"tool":33,"command":34},"npx","npx skills add Truc350\u002FWebHomeStay --skill google-adk-python",{"tool":36,"command":37},"bunx","bunx skills add Truc350\u002FWebHomeStay --skill google-adk-python",{"tool":39,"command":40},"pnpm","pnpm dlx skills add Truc350\u002FWebHomeStay --skill google-adk-python","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002FTruc350\u002FWebHomeStay\u002Fblob\u002Fmain\u002F.agents\u002Fskills\u002Fgoogle-adk-python\u002FSKILL.md",true,[],[45,52,59,66,74,81],{"skillNo":46,"name":47,"nameZh":6,"description":48,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":49,"categories":50,"keywords":51,"publishedAt":19},"01M7C3D9JFFHDYQ93YQ2NFSTXV","widget-cms","Tạo, sửa và kiểm thử widget Laravel CMS theo renderer theme\u002Fcore\u002Fplugin, layout builder, settings_schema, shortcode, hook, CSS\u002FJS và zone contract.","https:\u002F\u002Fgithub.com\u002FTruc350\u002FWebHomeStay\u002Fblob\u002Fmain\u002F.agents\u002Fskills\u002Fwidget-cms\u002FSKILL.md",[],[],{"skillNo":53,"name":54,"nameZh":6,"description":55,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":56,"categories":57,"keywords":58,"publishedAt":19},"01M7C3CNZVN589PK7ZJ1JZ98JB","website-sitemap","This skill should be used to create, crawl, map, redesign, or migrate website sitemaps and information architecture from briefs, ERP data, URLs, documents, or mixed inputs.","https:\u002F\u002Fgithub.com\u002FTruc350\u002FWebHomeStay\u002Fblob\u002Fmain\u002F.agents\u002Fskills\u002Fwebsite-sitemap\u002FSKILL.md",[],[],{"skillNo":60,"name":61,"nameZh":6,"description":62,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":63,"categories":64,"keywords":65,"publishedAt":19},"01M7C3C36PECBZQR0RBXGE8781","web-frameworks","Build with Next.js (App Router, RSC, SSR, ISR), Turborepo monorepos. Use for React apps, server rendering, build optimization, caching strategies, shared dependencies.","https:\u002F\u002Fgithub.com\u002FTruc350\u002FWebHomeStay\u002Fblob\u002Fmain\u002F.agents\u002Fskills\u002Fweb-frameworks\u002FSKILL.md",[],[],{"skillNo":67,"name":68,"nameZh":6,"description":69,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":70,"sourceHtmlUrl":71,"categories":72,"keywords":73,"publishedAt":19},"01M7C3A7X0WG8RWDDJFB9CHP0F","ui-styling","Style UIs with shadcn\u002Fui components (Radix UI + Tailwind CSS). Use for accessible components, themes, dark mode, responsive layouts, design systems, color customization.","https:\u002F\u002Fgithub.com\u002FTruc350\u002FWebHomeStay\u002Fblob\u002Fmain\u002F.agents\u002Fskills\u002Fui-styling\u002FLICENSE.txt","https:\u002F\u002Fgithub.com\u002FTruc350\u002FWebHomeStay\u002Fblob\u002Fmain\u002F.agents\u002Fskills\u002Fui-styling\u002FSKILL.md",[],[],{"skillNo":75,"name":76,"nameZh":6,"description":77,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":78,"categories":79,"keywords":80,"publishedAt":19},"01M7C39497ER4DVF3R8P9R6K41","threejs","Build 3D web apps with Three.js (WebGL\u002FWebGPU). 556 searchable examples, 60 API classes, 20 use cases. Actions: create 3D scene, load model, add animation, implement physics, build VR\u002FXR. Topics: GLTF loader, PBR materials, particle effects, shadows, post-processing, compute shaders, TSL. Integrations: WebGPU, physics engines, spatial audio.","https:\u002F\u002Fgithub.com\u002FTruc350\u002FWebHomeStay\u002Fblob\u002Fmain\u002F.agents\u002Fskills\u002Fthreejs\u002FSKILL.md",[],[],{"skillNo":82,"name":83,"nameZh":6,"description":84,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":85,"categories":86,"keywords":87,"publishedAt":19},"01M7C38TQE66MV142R5K58QJM6","theme-cms-ecommerce","This skill should be used independently to create, update, validate, or package complete Laravel CMS storefronts with product catalogs, search, promotions, cart, checkout, and responsive shopping UX.","https:\u002F\u002Fgithub.com\u002FTruc350\u002FWebHomeStay\u002Fblob\u002Fmain\u002F.agents\u002Fskills\u002Ftheme-cms-ecommerce\u002FSKILL.md",[],[]]