[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M7BVFVHSP1T18P59Y96RB77M":3,"skill-same:01M7BVFVHSP1T18P59Y96RB77M":79},{"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":51,"commitSha":52,"skillDirectory":53,"forksCount":54,"repositoryHtmlUrl":55,"repositoryHomepageUrl":6,"ownerHtmlUrl":56,"ownerBlogUrl":6,"ownerDescription":57,"updatedAt":58,"install":59,"resources":72},"01M7BVFVHSP1T18P59Y96RB77M","ai-agent-builder",null,"Build AI agents with tools, memory, and multi-step reasoning - ChatGPT, Claude, Gemini integration patterns","om-scogo\u002Fskillsh-scraper","om-scogo","https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F160466840?v=4","01M79AWB6DFVCQ7RK70CS34B96","Om Achrekar","01M79AWB6ZM9SKR8KDH1WSSM22",0,"MIT","https:\u002F\u002Fgithub.com\u002Fom-scogo\u002Fskillsh-scraper\u002Fblob\u002Fmain\u002Fdata\u002Fskills\u002Fclaude-office-skills-skills\u002Fai-agent-builder\u002FSkill.md",[],[],"2026-11-15 14:19:17",false,"---\nname: ai-agent-builder\ndescription: \"Build AI agents with tools, memory, and multi-step reasoning - ChatGPT, Claude, Gemini integration patterns\"\nversion: \"1.0.0\"\nauthor: claude-office-skills\nlicense: MIT\n\ncategory: ai\ntags:\n  - ai-agent\n  - chatgpt\n  - openai\n  - langchain\n  - automation\ndepartment: Engineering\n\nmodels:\n  recommended:\n    - claude-opus-4\n    - claude-sonnet-4\n\ncapabilities:\n  - agent_design\n  - tool_integration\n  - memory_management\n  - multi_step_reasoning\n  - conversation_flow\n\nlanguages:\n  - en\n  - zh\n\nrelated_skills:\n  - deep-research\n  - n8n-workflow\n  - slack-workflows\n---\n\n# AI Agent Builder\n\nDesign and build AI agents with tools, memory, and multi-step reasoning capabilities. Covers ChatGPT, Claude, Gemini integration patterns based on n8n's 5,000+ AI workflow templates.\n\n## Overview\n\nThis skill covers:\n- AI agent architecture design\n- Tool\u002Ffunction calling patterns\n- Memory and context management\n- Multi-step reasoning workflows\n- Platform integrations (Slack, Telegram, Web)\n\n---\n\n## AI Agent Architecture\n\n### Core Components\n\n```\n┌─────────────────────────────────────────────────────────────────┐\n│                      AI AGENT ARCHITECTURE                       │\n├─────────────────────────────────────────────────────────────────┤\n│                                                                 │\n│  ┌─────────────┐     ┌─────────────┐     ┌─────────────┐       │\n│  │   Input     │────▶│   Agent     │────▶│   Output    │       │\n│  │  (Query)    │     │   (LLM)     │     │  (Response) │       │\n│  └─────────────┘     └──────┬──────┘     └─────────────┘       │\n│                             │                                   │\n│         ┌───────────────────┼───────────────────┐              │\n│         │                   │                   │              │\n│         ▼                   ▼                   ▼              │\n│  ┌─────────────┐     ┌─────────────┐     ┌─────────────┐       │\n│  │   Tools     │     │   Memory    │     │  Knowledge  │       │\n│  │ (Functions) │     │  (Context)  │     │   (RAG)     │       │\n│  └─────────────┘     └─────────────┘     └─────────────┘       │\n│                                                                 │\n└─────────────────────────────────────────────────────────────────┘\n```\n\n### Agent Types\n\n```yaml\nagent_types:\n  reactive_agent:\n    description: \"Single-turn response, no memory\"\n    use_case: simple_qa, classification\n    complexity: low\n    \n  conversational_agent:\n    description: \"Multi-turn with conversation memory\"\n    use_case: chatbots, support\n    complexity: medium\n    \n  tool_using_agent:\n    description: \"Can call external tools\u002FAPIs\"\n    use_case: data_lookup, actions\n    complexity: medium\n    \n  reasoning_agent:\n    description: \"Multi-step planning and execution\"\n    use_case: complex_tasks, research\n    complexity: high\n    \n  multi_agent:\n    description: \"Multiple specialized agents collaborating\"\n    use_case: complex_workflows\n    complexity: very_high\n```\n\n---\n\n## Tool Calling Pattern\n\n### Tool Definition\n\n```yaml\ntool_definition:\n  name: \"get_weather\"\n  description: \"Get current weather for a location\"\n  parameters:\n    type: object\n    properties:\n      location:\n        type: string\n        description: \"City name or coordinates\"\n      units:\n        type: string\n        enum: [\"celsius\", \"fahrenheit\"]\n        default: \"celsius\"\n    required: [\"location\"]\n    \n  implementation:\n    type: api_call\n    endpoint: \"https:\u002F\u002Fapi.weather.com\u002Fv1\u002Fcurrent\"\n    method: GET\n    params:\n      q: \"{location}\"\n      units: \"{units}\"\n```\n\n### Common Tool Categories\n\n```yaml\ntool_categories:\n  data_retrieval:\n    - web_search: search the internet\n    - database_query: query SQL\u002FNoSQL\n    - api_lookup: call external APIs\n    - file_read: read documents\n    \n  actions:\n    - send_email: send emails\n    - create_calendar: schedule events\n    - update_crm: modify CRM records\n    - post_slack: send Slack messages\n    \n  computation:\n    - calculator: math operations\n    - code_interpreter: run Python\n    - data_analysis: analyze datasets\n    \n  generation:\n    - image_generation: create images\n    - document_creation: generate docs\n    - chart_creation: create visualizations\n```\n\n### n8n Tool Integration\n\n```yaml\nn8n_agent_workflow:\n  nodes:\n    - trigger:\n        type: webhook\n        path: \"\u002Fai-agent\"\n        \n    - ai_agent:\n        type: \"@n8n\u002Fn8n-nodes-langchain.agent\"\n        model: openai_gpt4\n        system_prompt: |\n          You are a helpful assistant that can:\n          1. Search the web for information\n          2. Query our customer database\n          3. Send emails on behalf of the user\n          \n        tools:\n          - web_search\n          - database_query\n          - send_email\n          \n    - respond:\n        type: respond_to_webhook\n        data: \"{{ $json.output }}\"\n```\n\n---\n\n## Memory Patterns\n\n### Memory Types\n\n```yaml\nmemory_types:\n  buffer_memory:\n    description: \"Store last N messages\"\n    implementation: |\n      messages = []\n      def add_message(role, content):\n          messages.append({\"role\": role, \"content\": content})\n          if len(messages) > MAX_MESSAGES:\n              messages.pop(0)\n    use_case: simple_chatbots\n    \n  summary_memory:\n    description: \"Summarize conversation periodically\"\n    implementation: |\n      When messages > threshold:\n          summary = llm.summarize(messages[:-5])\n          messages = [summary_message] + messages[-5:]\n    use_case: long_conversations\n    \n  vector_memory:\n    description: \"Store in vector DB for semantic retrieval\"\n    implementation: |\n      # Store\n      embedding = embed(message)\n      vector_db.insert(embedding, message)\n      \n      # Retrieve\n      relevant = vector_db.search(query_embedding, k=5)\n    use_case: knowledge_retrieval\n    \n  entity_memory:\n    description: \"Track entities mentioned in conversation\"\n    implementation: |\n      entities = {}\n      def update_entities(message):\n          extracted = llm.extract_entities(message)\n          entities.update(extracted)\n    use_case: personalized_assistants\n```\n\n### Context Window Management\n\n```yaml\ncontext_management:\n  strategies:\n    sliding_window:\n      keep: last_n_messages\n      n: 10\n      \n    relevance_based:\n      method: embed_and_rank\n      keep: top_k_relevant\n      k: 5\n      \n    hierarchical:\n      levels:\n        - immediate: last_3_messages\n        - recent: summary_of_last_10\n        - long_term: key_facts_from_all\n        \n  token_budget:\n    total: 8000\n    system_prompt: 1000\n    tools: 1000\n    memory: 4000\n    current_query: 1000\n    response: 1000\n```\n\n---\n\n## Multi-Step Reasoning\n\n### ReAct Pattern\n\n```\nThought: I need to find information about X\nAction: web_search(\"X\")\nObservation: [search results]\nThought: Based on the results, I should also check Y\nAction: database_query(\"SELECT * FROM Y\")\nObservation: [database results]\nThought: Now I have enough information to answer\nAction: respond(\"Final answer based on X and Y\")\n```\n\n### Planning Agent\n\n```yaml\nplanning_workflow:\n  step_1_plan:\n    prompt: |\n      Task: {user_request}\n      \n      Create a step-by-step plan to complete this task.\n      Each step should be specific and actionable.\n      \n    output: numbered_steps\n    \n  step_2_execute:\n    for_each: step\n    actions:\n      - execute_step\n      - validate_result\n      - adjust_if_needed\n      \n  step_3_synthesize:\n    prompt: |\n      Steps completed: {executed_steps}\n      Results: {results}\n      \n      Synthesize a final response for the user.\n```\n\n---\n\n## Platform Integrations\n\n### Slack Bot Agent\n\n```yaml\nslack_agent:\n  trigger: slack_message\n  \n  workflow:\n    1. receive_message:\n        extract: [user, channel, text, thread_ts]\n        \n    2. get_context:\n        if: thread_ts\n        action: fetch_thread_history\n        \n    3. process_with_agent:\n        model: gpt-4\n        system: \"You are a helpful Slack assistant\"\n        tools: [web_search, jira_lookup, calendar_check]\n        \n    4. respond:\n        action: post_to_slack\n        channel: \"{channel}\"\n        thread_ts: \"{thread_ts}\"\n        text: \"{agent_response}\"\n```\n\n### Telegram Bot Agent\n\n```yaml\ntelegram_agent:\n  trigger: telegram_message\n  \n  handlers:\n    text_message:\n      - extract_text\n      - process_with_ai\n      - send_response\n      \n    voice_message:\n      - transcribe_with_whisper\n      - process_with_ai\n      - send_text_or_voice_response\n      \n    image:\n      - analyze_with_vision\n      - process_with_ai\n      - send_response\n      \n    document:\n      - extract_content\n      - process_with_ai\n      - send_response\n```\n\n### Web Chat Interface\n\n```yaml\nweb_chat_agent:\n  frontend:\n    type: react_component\n    features:\n      - message_input\n      - message_history\n      - typing_indicator\n      - file_upload\n      \n  backend:\n    endpoint: \"\u002Fapi\u002Fchat\"\n    method: POST\n    streaming: true\n    \n  session_management:\n    method: jwt_token\n    storage: redis\n    ttl: 24_hours\n```\n\n---\n\n## Agent Templates\n\n### Customer Support Agent\n\n```yaml\nsupport_agent:\n  name: \"Customer Support AI\"\n  model: gpt-4\n  \n  system_prompt: |\n    You are a customer support agent for {company}.\n    \n    Guidelines:\n    - Be helpful, professional, and empathetic\n    - Use the knowledge base to answer questions\n    - Escalate to human if you cannot help\n    - Never make up information\n    \n    Available actions:\n    - Search knowledge base\n    - Look up customer account\n    - Create support ticket\n    - Escalate to human agent\n    \n  tools:\n    - knowledge_search:\n        description: \"Search help articles\"\n    - customer_lookup:\n        description: \"Get customer account info\"\n    - create_ticket:\n        description: \"Create support ticket\"\n    - escalate:\n        description: \"Transfer to human agent\"\n        \n  escalation_triggers:\n    - sentiment: very_negative\n    - keywords: [\"speak to human\", \"manager\", \"complaint\"]\n    - failed_resolution: after_3_attempts\n```\n\n### Research Agent\n\n```yaml\nresearch_agent:\n  name: \"Research Assistant\"\n  model: claude-opus\n  \n  system_prompt: |\n    You are a research assistant that helps users find \n    and synthesize information on any topic.\n    \n    Process:\n    1. Understand the research question\n    2. Plan search strategy\n    3. Execute searches across multiple sources\n    4. Synthesize findings\n    5. Present with citations\n    \n  tools:\n    - web_search\n    - academic_search\n    - news_search\n    - wikipedia_lookup\n    \n  output_format:\n    - executive_summary\n    - key_findings\n    - detailed_analysis\n    - sources_cited\n```\n\n---\n\n## Output Example\n\n**Request**: \"Design an AI agent for appointment scheduling\"\n\n**Output**:\n\n```markdown\n# Appointment Scheduling Agent Design\n\n## Architecture\n\n```yaml\nagent:\n  name: \"Scheduling Assistant\"\n  model: gpt-4\n  \n  capabilities:\n    - Check calendar availability\n    - Book appointments\n    - Send confirmations\n    - Handle rescheduling\n    - Answer FAQs\n```\n\n## Tools Required\n\n1. **calendar_check**\n   - Input: date_range, duration\n   - Output: available_slots[]\n\n2. **book_appointment**\n   - Input: datetime, customer_info, service_type\n   - Output: confirmation_id\n\n3. **send_notification**\n   - Input: type, recipient, details\n   - Output: success\u002Ffailure\n\n4. **customer_lookup**\n   - Input: email or phone\n   - Output: customer_record\n\n## Conversation Flow\n\n```\nUser: \"I'd like to book an appointment\"\nAgent: [Thought: Need to know service type and preferred time]\nAgent: \"I'd be happy to help! What service are you looking for, \n        and do you have a preferred date\u002Ftime?\"\n\nUser: \"Haircut, next Tuesday afternoon\"\nAgent: [Action: calendar_check(date=\"next Tuesday\", time=\"afternoon\")]\nAgent: \"I have availability at 2pm, 3pm, and 4:30pm next Tuesday. \n        Which works best for you?\"\n\nUser: \"3pm please\"\nAgent: [Action: book_appointment(datetime=\"Tuesday 3pm\", service=\"haircut\")]\nAgent: [Action: send_notification(type=\"confirmation\", ...)]\nAgent: \"You're all set for a haircut at 3pm next Tuesday. \n        I've sent a confirmation to your email. See you then!\"\n```\n\n## n8n Implementation\n\n```yaml\nworkflow:\n  - webhook_trigger: \u002Fschedule-chat\n  - ai_agent:\n      tools: [calendar, booking, notification]\n  - respond_to_user\n```\n```\n\n---\n\n*AI Agent Builder Skill - Part of Claude Office Skills*\n","# AI Agent Builder\n\nDesign and build AI agents with tools, memory, and multi-step reasoning capabilities. Covers ChatGPT, Claude, Gemini integration patterns based on n8n's 5,000+ AI workflow templates.\n\n## Overview\n\nThis skill covers:\n- AI agent architecture design\n- Tool\u002Ffunction calling patterns\n- Memory and context management\n- Multi-step reasoning workflows\n- Platform integrations (Slack, Telegram, Web)\n\n---\n\n## AI Agent Architecture\n\n### Core Components\n\n```\n┌─────────────────────────────────────────────────────────────────┐\n│                      AI AGENT ARCHITECTURE                       │\n├─────────────────────────────────────────────────────────────────┤\n│                                                                 │\n│  ┌─────────────┐     ┌─────────────┐     ┌─────────────┐       │\n│  │   Input     │────▶│   Agent     │────▶│   Output    │       │\n│  │  (Query)    │     │   (LLM)     │     │  (Response) │       │\n│  └─────────────┘     └──────┬──────┘     └─────────────┘       │\n│                             │                                   │\n│         ┌───────────────────┼───────────────────┐              │\n│         │                   │                   │              │\n│         ▼                   ▼                   ▼              │\n│  ┌─────────────┐     ┌─────────────┐     ┌─────────────┐       │\n│  │   Tools     │     │   Memory    │     │  Knowledge  │       │\n│  │ (Functions) │     │  (Context)  │     │   (RAG)     │       │\n│  └─────────────┘     └─────────────┘     └─────────────┘       │\n│                                                                 │\n└─────────────────────────────────────────────────────────────────┘\n```\n\n### Agent Types\n\n```yaml\nagent_types:\n  reactive_agent:\n    description: \"Single-turn response, no memory\"\n    use_case: simple_qa, classification\n    complexity: low\n    \n  conversational_agent:\n    description: \"Multi-turn with conversation memory\"\n    use_case: chatbots, support\n    complexity: medium\n    \n  tool_using_agent:\n    description: \"Can call external tools\u002FAPIs\"\n    use_case: data_lookup, actions\n    complexity: medium\n    \n  reasoning_agent:\n    description: \"Multi-step planning and execution\"\n    use_case: complex_tasks, research\n    complexity: high\n    \n  multi_agent:\n    description: \"Multiple specialized agents collaborating\"\n    use_case: complex_workflows\n    complexity: very_high\n```\n\n---\n\n## Tool Calling Pattern\n\n### Tool Definition\n\n```yaml\ntool_definition:\n  name: \"get_weather\"\n  description: \"Get current weather for a location\"\n  parameters:\n    type: object\n    properties:\n      location:\n        type: string\n        description: \"City name or coordinates\"\n      units:\n        type: string\n        enum: [\"celsius\", \"fahrenheit\"]\n        default: \"celsius\"\n    required: [\"location\"]\n    \n  implementation:\n    type: api_call\n    endpoint: \"https:\u002F\u002Fapi.weather.com\u002Fv1\u002Fcurrent\"\n    method: GET\n    params:\n      q: \"{location}\"\n      units: \"{units}\"\n```\n\n### Common Tool Categories\n\n```yaml\ntool_categories:\n  data_retrieval:\n    - web_search: search the internet\n    - database_query: query SQL\u002FNoSQL\n    - api_lookup: call external APIs\n    - file_read: read documents\n    \n  actions:\n    - send_email: send emails\n    - create_calendar: schedule events\n    - update_crm: modify CRM records\n    - post_slack: send Slack messages\n    \n  computation:\n    - calculator: math operations\n    - code_interpreter: run Python\n    - data_analysis: analyze datasets\n    \n  generation:\n    - image_generation: create images\n    - document_creation: generate docs\n    - chart_creation: create visualizations\n```\n\n### n8n Tool Integration\n\n```yaml\nn8n_agent_workflow:\n  nodes:\n    - trigger:\n        type: webhook\n        path: \"\u002Fai-agent\"\n        \n    - ai_agent:\n        type: \"@n8n\u002Fn8n-nodes-langchain.agent\"\n        model: openai_gpt4\n        system_prompt: |\n          You are a helpful assistant that can:\n          1. Search the web for information\n          2. Query our customer database\n          3. Send emails on behalf of the user\n          \n        tools:\n          - web_search\n          - database_query\n          - send_email\n          \n    - respond:\n        type: respond_to_webhook\n        data: \"{{ $json.output }}\"\n```\n\n---\n\n## Memory Patterns\n\n### Memory Types\n\n```yaml\nmemory_types:\n  buffer_memory:\n    description: \"Store last N messages\"\n    implementation: |\n      messages = []\n      def add_message(role, content):\n          messages.append({\"role\": role, \"content\": content})\n          if len(messages) > MAX_MESSAGES:\n              messages.pop(0)\n    use_case: simple_chatbots\n    \n  summary_memory:\n    description: \"Summarize conversation periodically\"\n    implementation: |\n      When messages > threshold:\n          summary = llm.summarize(messages[:-5])\n          messages = [summary_message] + messages[-5:]\n    use_case: long_conversations\n    \n  vector_memory:\n    description: \"Store in vector DB for semantic retrieval\"\n    implementation: |\n      # Store\n      embedding = embed(message)\n      vector_db.insert(embedding, message)\n      \n      # Retrieve\n      relevant = vector_db.search(query_embedding, k=5)\n    use_case: knowledge_retrieval\n    \n  entity_memory:\n    description: \"Track entities mentioned in conversation\"\n    implementation: |\n      entities = {}\n      def update_entities(message):\n          extracted = llm.extract_entities(message)\n          entities.update(extracted)\n    use_case: personalized_assistants\n```\n\n### Context Window Management\n\n```yaml\ncontext_management:\n  strategies:\n    sliding_window:\n      keep: last_n_messages\n      n: 10\n      \n    relevance_based:\n      method: embed_and_rank\n      keep: top_k_relevant\n      k: 5\n      \n    hierarchical:\n      levels:\n        - immediate: last_3_messages\n        - recent: summary_of_last_10\n        - long_term: key_facts_from_all\n        \n  token_budget:\n    total: 8000\n    system_prompt: 1000\n    tools: 1000\n    memory: 4000\n    current_query: 1000\n    response: 1000\n```\n\n---\n\n## Multi-Step Reasoning\n\n### ReAct Pattern\n\n```\nThought: I need to find information about X\nAction: web_search(\"X\")\nObservation: [search results]\nThought: Based on the results, I should also check Y\nAction: database_query(\"SELECT * FROM Y\")\nObservation: [database results]\nThought: Now I have enough information to answer\nAction: respond(\"Final answer based on X and Y\")\n```\n\n### Planning Agent\n\n```yaml\nplanning_workflow:\n  step_1_plan:\n    prompt: |\n      Task: {user_request}\n      \n      Create a step-by-step plan to complete this task.\n      Each step should be specific and actionable.\n      \n    output: numbered_steps\n    \n  step_2_execute:\n    for_each: step\n    actions:\n      - execute_step\n      - validate_result\n      - adjust_if_needed\n      \n  step_3_synthesize:\n    prompt: |\n      Steps completed: {executed_steps}\n      Results: {results}\n      \n      Synthesize a final response for the user.\n```\n\n---\n\n## Platform Integrations\n\n### Slack Bot Agent\n\n```yaml\nslack_agent:\n  trigger: slack_message\n  \n  workflow:\n    1. receive_message:\n        extract: [user, channel, text, thread_ts]\n        \n    2. get_context:\n        if: thread_ts\n        action: fetch_thread_history\n        \n    3. process_with_agent:\n        model: gpt-4\n        system: \"You are a helpful Slack assistant\"\n        tools: [web_search, jira_lookup, calendar_check]\n        \n    4. respond:\n        action: post_to_slack\n        channel: \"{channel}\"\n        thread_ts: \"{thread_ts}\"\n        text: \"{agent_response}\"\n```\n\n### Telegram Bot Agent\n\n```yaml\ntelegram_agent:\n  trigger: telegram_message\n  \n  handlers:\n    text_message:\n      - extract_text\n      - process_with_ai\n      - send_response\n      \n    voice_message:\n      - transcribe_with_whisper\n      - process_with_ai\n      - send_text_or_voice_response\n      \n    image:\n      - analyze_with_vision\n      - process_with_ai\n      - send_response\n      \n    document:\n      - extract_content\n      - process_with_ai\n      - send_response\n```\n\n### Web Chat Interface\n\n```yaml\nweb_chat_agent:\n  frontend:\n    type: react_component\n    features:\n      - message_input\n      - message_history\n      - typing_indicator\n      - file_upload\n      \n  backend:\n    endpoint: \"\u002Fapi\u002Fchat\"\n    method: POST\n    streaming: true\n    \n  session_management:\n    method: jwt_token\n    storage: redis\n    ttl: 24_hours\n```\n\n---\n\n## Agent Templates\n\n### Customer Support Agent\n\n```yaml\nsupport_agent:\n  name: \"Customer Support AI\"\n  model: gpt-4\n  \n  system_prompt: |\n    You are a customer support agent for {company}.\n    \n    Guidelines:\n    - Be helpful, professional, and empathetic\n    - Use the knowledge base to answer questions\n    - Escalate to human if you cannot help\n    - Never make up information\n    \n    Available actions:\n    - Search knowledge base\n    - Look up customer account\n    - Create support ticket\n    - Escalate to human agent\n    \n  tools:\n    - knowledge_search:\n        description: \"Search help articles\"\n    - customer_lookup:\n        description: \"Get customer account info\"\n    - create_ticket:\n        description: \"Create support ticket\"\n    - escalate:\n        description: \"Transfer to human agent\"\n        \n  escalation_triggers:\n    - sentiment: very_negative\n    - keywords: [\"speak to human\", \"manager\", \"complaint\"]\n    - failed_resolution: after_3_attempts\n```\n\n### Research Agent\n\n```yaml\nresearch_agent:\n  name: \"Research Assistant\"\n  model: claude-opus\n  \n  system_prompt: |\n    You are a research assistant that helps users find \n    and synthesize information on any topic.\n    \n    Process:\n    1. Understand the research question\n    2. Plan search strategy\n    3. Execute searches across multiple sources\n    4. Synthesize findings\n    5. Present with citations\n    \n  tools:\n    - web_search\n    - academic_search\n    - news_search\n    - wikipedia_lookup\n    \n  output_format:\n    - executive_summary\n    - key_findings\n    - detailed_analysis\n    - sources_cited\n```\n\n---\n\n## Output Example\n\n**Request**: \"Design an AI agent for appointment scheduling\"\n\n**Output**:\n\n```markdown\n# Appointment Scheduling Agent Design\n\n## Architecture\n\n```yaml\nagent:\n  name: \"Scheduling Assistant\"\n  model: gpt-4\n  \n  capabilities:\n    - Check calendar availability\n    - Book appointments\n    - Send confirmations\n    - Handle rescheduling\n    - Answer FAQs\n```\n\n## Tools Required\n\n1. **calendar_check**\n   - Input: date_range, duration\n   - Output: available_slots[]\n\n2. **book_appointment**\n   - Input: datetime, customer_info, service_type\n   - Output: confirmation_id\n\n3. **send_notification**\n   - Input: type, recipient, details\n   - Output: success\u002Ffailure\n\n4. **customer_lookup**\n   - Input: email or phone\n   - Output: customer_record\n\n## Conversation Flow\n\n```\nUser: \"I'd like to book an appointment\"\nAgent: [Thought: Need to know service type and preferred time]\nAgent: \"I'd be happy to help! What service are you looking for, \n        and do you have a preferred date\u002Ftime?\"\n\nUser: \"Haircut, next Tuesday afternoon\"\nAgent: [Action: calendar_check(date=\"next Tuesday\", time=\"afternoon\")]\nAgent: \"I have availability at 2pm, 3pm, and 4:30pm next Tuesday. \n        Which works best for you?\"\n\nUser: \"3pm please\"\nAgent: [Action: book_appointment(datetime=\"Tuesday 3pm\", service=\"haircut\")]\nAgent: [Action: send_notification(type=\"confirmation\", ...)]\nAgent: \"You're all set for a haircut at 3pm next Tuesday. \n        I've sent a confirmation to your email. See you then!\"\n```\n\n## n8n Implementation\n\n```yaml\nworkflow:\n  - webhook_trigger: \u002Fschedule-chat\n  - ai_agent:\n      tools: [calendar, booking, notification]\n  - respond_to_user\n```\n```\n\n---\n\n*AI Agent Builder Skill - Part of Claude Office Skills*\n",{"name":5,"tags":24,"author":30,"models":31,"license":15,"version":35,"category":36,"languages":37,"department":40,"description":7,"capabilities":41,"related_skills":47},[25,26,27,28,29],"ai-agent","chatgpt","openai","langchain","automation","claude-office-skills",{"recommended":32},[33,34],"claude-opus-4","claude-sonnet-4","1.0.0","ai",[38,39],"en","zh","Engineering",[42,43,44,45,46],"agent_design","tool_integration","memory_management","multi_step_reasoning","conversation_flow",[48,49,50],"deep-research","n8n-workflow","slack-workflows","main","8859218955d9504cd18b59e7892100718039d3b4","data\u002Fskills\u002Fclaude-office-skills-skills\u002Fai-agent-builder",1,"https:\u002F\u002Fgithub.com\u002Fom-scogo\u002Fskillsh-scraper","https:\u002F\u002Fgithub.com\u002Fom-scogo","SDE-1 @ Scogo ","2026-07-18 12:22:46",{"source":8,"skillName":5,"commands":60,"prompt":70,"zipAvailable":71},[61,64,67],{"tool":62,"command":63},"npx","npx skills add om-scogo\u002Fskillsh-scraper --skill ai-agent-builder",{"tool":65,"command":66},"bunx","bunx skills add om-scogo\u002Fskillsh-scraper --skill ai-agent-builder",{"tool":68,"command":69},"pnpm","pnpm dlx skills add om-scogo\u002Fskillsh-scraper --skill ai-agent-builder","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002Fom-scogo\u002Fskillsh-scraper\u002Fblob\u002Fmain\u002Fdata\u002Fskills\u002Fclaude-office-skills-skills\u002Fai-agent-builder\u002FSkill.md",true,[73],{"path":74,"resourceType":75,"resourceTypeDesc":76,"sizeBytes":77,"sourceHtmlUrl":78},"metadata.json","REFERENCE","参考文档",484,"https:\u002F\u002Fgithub.com\u002Fom-scogo\u002Fskillsh-scraper\u002Fblob\u002Fmain\u002Fdata\u002Fskills\u002Fclaude-office-skills-skills\u002Fai-agent-builder\u002Fmetadata.json",[80,87,92,99,104,111],{"skillNo":81,"name":82,"nameZh":6,"description":83,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":84,"categories":85,"keywords":86,"publishedAt":19},"01M7D6MC9FAAEV4CMNYX8W7JTG","arxiv-search","Search arXiv physics, math, and computer science preprints using natural language queries. Powered by Valyu semantic search.","https:\u002F\u002Fgithub.com\u002Fom-scogo\u002Fskillsh-scraper\u002Fblob\u002Fmain\u002Fdata\u002Fskills\u002Fyorkeccak-scientific-skills\u002Farxiv-search\u002Fskills\u002Farxiv-search\u002FSKILL.md",[],[],{"skillNo":88,"name":82,"nameZh":6,"description":83,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":89,"categories":90,"keywords":91,"publishedAt":19},"01M7D6M85CKKN87TA98VCQV4H7","https:\u002F\u002Fgithub.com\u002Fom-scogo\u002Fskillsh-scraper\u002Fblob\u002Fmain\u002Fdata\u002Fskills\u002Fyorkeccak-scientific-skills\u002Farxiv-search\u002FSkill.md",[],[],{"skillNo":93,"name":94,"nameZh":6,"description":95,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":96,"categories":97,"keywords":98,"publishedAt":19},"01M7D4ZTCS4DR3CE4JY06X84T2","working-with-legacy-code","Safely change and test untested codebases using Feathers' \"Working Effectively with Legacy Code\". Use when the user mentions \"legacy code\", \"no tests\", \"untested codebase\", \"how do I test this\", \"seams\", \"characterization tests\", \"golden master\", \"sprout method\", \"afraid to change this code\", \"monster method\", \"dependency breaking\", or \"inherited a messy codebase\". Also trigger when changing code without tests safely, getting a class under test when constructors, statics, or singletons block it, adding features to tangled modules, or planning incremental test coverage for an old codebase. Covers the legacy-code change algorithm, seams, characterization tests, sprout\u002Fwrap, and dependency-breaking techniques. For refactoring code that already has tests, see refactoring-patterns. For day-to-day code quality, see clean-code.","https:\u002F\u002Fgithub.com\u002Fom-scogo\u002Fskillsh-scraper\u002Fblob\u002Fmain\u002Fdata\u002Fskills\u002Fwondelai-skills\u002Fworking-with-legacy-code\u002Fworking-with-legacy-code\u002FSKILL.md",[],[],{"skillNo":100,"name":94,"nameZh":6,"description":95,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":101,"categories":102,"keywords":103,"publishedAt":19},"01M7D4ZHZKJD2RSNGAF63V14P4","https:\u002F\u002Fgithub.com\u002Fom-scogo\u002Fskillsh-scraper\u002Fblob\u002Fmain\u002Fdata\u002Fskills\u002Fwondelai-skills\u002Fworking-with-legacy-code\u002FSkill.md",[],[],{"skillNo":105,"name":106,"nameZh":6,"description":107,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":108,"categories":109,"keywords":110,"publishedAt":19},"01M7D4ZB4CJBFW5JE67CPH2QRV","web-typography","Select, pair, and implement typefaces for web projects. Use when the user mentions \"font pairing\", \"which typeface\", \"line height\", \"responsive typography\", \"web font loading\", \"type hierarchy\", \"variable fonts\", \"FOUT\u002FFOIT\", \"typographic scale\", or \"the text is hard to read\". Also trigger when choosing between system fonts and web fonts, optimizing font-loading performance, or designing readable long-form content. Covers readability evaluation, CSS implementation, and performance optimization. For overall UI design systems, see refactoring-ui. For dramatic typographic experiences, see top-design.","https:\u002F\u002Fgithub.com\u002Fom-scogo\u002Fskillsh-scraper\u002Fblob\u002Fmain\u002Fdata\u002Fskills\u002Fwondelai-skills\u002Fweb-typography\u002Fweb-typography\u002FSKILL.md",[],[],{"skillNo":112,"name":106,"nameZh":6,"description":107,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":113,"categories":114,"keywords":115,"publishedAt":19},"01M7D4Z4MVYVY3NMJ18BNV6WNA","https:\u002F\u002Fgithub.com\u002Fom-scogo\u002Fskillsh-scraper\u002Fblob\u002Fmain\u002Fdata\u002Fskills\u002Fwondelai-skills\u002Fweb-typography\u002FSkill.md",[],[]]