botrun-flow-lang 5.12.262__py3-none-any.whl → 5.12.264__py3-none-any.whl

botrun_flow_lang/mcp_server/default_mcp.py

@@ -1,744 +1,744 @@
-import os
-import asyncio
-import json
-import pytz
-from datetime import datetime
-from typing import Dict, Any, Optional
-
-# Import for web_search
-from botrun_flow_lang.langgraph_agents.agents.util.perplexity_search import (
-    respond_with_perplexity_search,
-)
-
-from mcp.server.fastmcp import FastMCP
-from langchain_core.runnables import RunnableConfig
-
-# Import necessary dependencies
-from botrun_flow_lang.models.nodes.utils import scrape_single_url
-from botrun_flow_lang.langgraph_agents.agents.util.pdf_analyzer import (
-    analyze_pdf_async,
-)
-from botrun_flow_lang.langgraph_agents.agents.util.img_util import analyze_imgs
-from botrun_flow_lang.langgraph_agents.agents.util.local_files import (
-    upload_and_get_tmp_public_url,
-)
-from botrun_flow_lang.langgraph_agents.agents.util.html_util import generate_html_file
-from botrun_flow_lang.langgraph_agents.agents.util.plotly_util import (
-    generate_plotly_files,
-)
-from botrun_flow_lang.langgraph_agents.agents.util.mermaid_util import (
-    generate_mermaid_files,
-)
-from botrun_flow_lang.utils.clients.rate_limit_client import RateLimitClient
-from botrun_flow_lang.utils.botrun_logger import get_default_botrun_logger
-
-# Import for generate_image
-from langchain_community.utilities.dalle_image_generator import DallEAPIWrapper
-from langchain_community.callbacks import get_openai_callback
-
-# Initialize MCP server
-mcp = FastMCP(name="BotrunFlowLangDefaultMCP", stateless_http=True)
-
-# Initialize logger - reuse the default instance
-logger = get_default_botrun_logger()
-
-
-# Exception class for rate limit errors
-class BotrunRateLimitException(Exception):
-    """Exception that should be displayed directly to the user."""
-
-    def __init__(self, message):
-        self.message = f"[Please tell user error] {message}"
-        super().__init__(self.message)
-
-
-@mcp.tool()
-async def scrape(url: str) -> dict:
-    """
-    Scrape a web page to extract its content.
-
-    Args:
-        url: The URL to scrape
-
-    Returns:
-        dict: A dictionary containing the scraped content or error message
-    """
-    try:
-        logger.info(f"scrape {url}")
-        return await scrape_single_url(url)
-    except Exception as e:
-        logger.error(f"scrape {url} error: {e}", error=str(e), exc_info=True)
-        return {"url": url, "status": "error", "error": str(e)}
-
-
-@mcp.tool()
-async def chat_with_pdf(
-    pdf_url: str, user_input: str, botrun_flow_lang_url: str, user_id: str
-) -> str:
-    """
-    Analyze a PDF file and answer questions about its content.
-
-    Supports intelligent processing based on file size:
-    - Small files (< 5MB): Direct multimodal analysis
-    - Large files (>= 5MB): Compress -> Split -> Parallel multimodal Q&A -> Merge results
-
-    Args:
-        pdf_url: The URL to the PDF file (can be generated using generate_tmp_public_url for local files)
-        user_input: The user's question or instruction about the PDF content
-        botrun_flow_lang_url: REQUIRED - URL for the botrun flow lang API (LLM can get this from system prompt)
-        user_id: REQUIRED - User ID for file upload (LLM can get this from system prompt)
-
-    Returns:
-        str: Analysis result or Plotly-compatible data structure if visualization is needed
-    """
-    logger.info(f"chat_with_pdf pdf_url: {pdf_url} user_input: {user_input}")
-
-    # Convert local file path to public URL if needed
-    if not pdf_url.startswith("http"):
-        pdf_url = upload_and_get_tmp_public_url(pdf_url, botrun_flow_lang_url, user_id)
-
-    return await analyze_pdf_async(pdf_url, user_input)
-
-
-@mcp.tool()
-async def chat_with_imgs(
-    img_urls: list[str],
-    user_input: str,
-    botrun_flow_lang_url: str,
-    user_id: str,
-) -> str:
-    """
-    Analyze multiple images and answer questions about their content.
-
-    Args:
-        img_urls: List of URLs to the image files (can be generated using generate_tmp_public_url for local files)
-        user_input: Question or instruction about the image content(s)
-        botrun_flow_lang_url: REQUIRED - URL for the botrun flow lang API (LLM can get this from system prompt)
-        user_id: REQUIRED - User ID for file upload (LLM can get this from system prompt)
-
-    Returns:
-        str: Analysis result or Plotly-compatible data structure if visualization is needed
-    """
-    logger.info(f"chat_with_imgs img_urls: {img_urls} user_input: {user_input}")
-
-    # Convert local file paths to public URLs if needed
-    new_img_urls = []
-    for img_url in img_urls:
-        if not img_url.startswith("http"):
-            img_url = upload_and_get_tmp_public_url(
-                img_url, botrun_flow_lang_url, user_id
-            )
-        new_img_urls.append(img_url)
-
-    return analyze_imgs(new_img_urls, user_input)
-
-
-@mcp.tool()
-async def generate_image(
-    user_input: str, user_id: str = "", botrun_flow_lang_url: str = ""
-) -> str:
-    """
-    Generate high-quality images using DALL-E 3 and store permanently in GCS.
-
-    When using generate_image tool, you must include the image URL in your response.
-    You MUST respond using this format (from @begin img to @end, including the image URL):
-    @begin img("{image_url}") @end
-
-    Capabilities:
-    - Creates photorealistic images and art
-    - Handles complex scenes and compositions
-    - Maintains consistent styles
-    - Follows detailed prompts with high accuracy
-    - Supports various artistic styles and mediums
-
-    Best practices for prompts:
-    - Be specific about style, mood, lighting, and composition
-    - Include details about perspective and setting
-    - Specify artistic medium if desired (e.g., "oil painting", "digital art")
-    - Mention color schemes or specific colors
-    - Describe the atmosphere or emotion you want to convey
-
-    Limitations:
-    - Cannot generate images of public figures or celebrities
-    - Avoids harmful, violent, or adult content
-    - May have inconsistencies with hands, faces, or text
-    - Cannot generate exact copies of existing artworks or brands
-    - Limited to single image generation per request
-    - Subject to daily usage limits
-
-    Args:
-        user_input: Detailed description of the image you want to generate.
-                   Be specific about style, content, and composition.
-        user_id: REQUIRED - User ID for rate limit and file storage (LLM can get this from system prompt)
-        botrun_flow_lang_url: REQUIRED - URL for the botrun flow lang API (LLM can get this from system prompt)
-
-    Returns:
-        str: Permanent URL to the generated image stored in GCS, or error message if generation fails
-    """
-    try:
-        logger.info(f"generate_image user_input: {user_input}")
-
-        # 驗證必要參數
-        if not user_id:
-            logger.error("User ID not available")
-            return "User ID not available"
-        if not botrun_flow_lang_url:
-            logger.error("botrun_flow_lang_url not available")
-            return "botrun_flow_lang_url not available"
-
-        # Check rate limit before generating image
-        rate_limit_client = RateLimitClient()
-        rate_limit_info = await rate_limit_client.get_rate_limit(user_id)
-
-        # Check if user can generate an image
-        drawing_info = rate_limit_info.get("drawing", {})
-        can_use = drawing_info.get("can_use", False)
-
-        if not can_use:
-            daily_limit = drawing_info.get("daily_limit", 0)
-            current_usage = drawing_info.get("current_usage", 0)
-            logger.error(
-                f"User {user_id} has reached daily limit of {daily_limit} image generations. "
-                f"Current usage: {current_usage}. Please try again tomorrow."
-            )
-            return f"[Please tell user error] You have reached your daily limit of {daily_limit} image generations. " \
-                f"Current usage: {current_usage}. Please try again tomorrow."
-            # raise BotrunRateLimitException(
-            #     f"You have reached your daily limit of {daily_limit} image generations. "
-            #     f"Current usage: {current_usage}. Please try again tomorrow."
-            # )
-
-        # 2. 使用 DALL-E 生成圖片
-        dalle_wrapper = DallEAPIWrapper(
-            api_key=os.getenv("OPENAI_API_KEY"), model="dall-e-3"
-        )
-
-        # Generate image with token usage tracking
-        with get_openai_callback() as cb:
-            temp_image_url = dalle_wrapper.run(user_input)
-            logger.info(
-                f"DALL-E generated temporary URL: {temp_image_url}, "
-                f"prompt tokens: {cb.prompt_tokens}, "
-                f"completion tokens: {cb.completion_tokens}"
-            )
-
-        # 3. 下載並上傳到 GCS，取得永久 URL
-        from botrun_flow_lang.langgraph_agents.agents.util.local_files import (
-            upload_image_and_get_public_url,
-        )
-
-        try:
-            permanent_url = await upload_image_and_get_public_url(
-                temp_image_url, botrun_flow_lang_url, user_id
-            )
-            logger.info(f"Image uploaded to GCS: {permanent_url}")
-
-            # 4. 更新使用計數
-            await rate_limit_client.update_drawing_usage(user_id)
-
-            return permanent_url
-        except Exception as upload_error:
-            logger.error(
-                f"Failed to upload to GCS, returning temporary URL: {upload_error}"
-            )
-            # Fallback: 回傳臨時 URL
-            await rate_limit_client.update_drawing_usage(user_id)
-            return temp_image_url
-
-    except Exception as e:
-        logger.error(f"generate_image error: {e}", error=str(e), exc_info=True)
-
-        # Check if this is a user-visible exception
-        if str(e).startswith("[Please tell user error]"):
-            return str(e)  # Return the error message as is
-        return f"Error: {e}"
-
-
-@mcp.tool()
-async def generate_tmp_public_url(
-    file_path: str, botrun_flow_lang_url: str, user_id: str
-) -> str:
-    """
-    Create a temporary public URL for a local file.
-
-    Args:
-        file_path: The path to the local file you want to make publicly accessible
-        botrun_flow_lang_url: REQUIRED - URL for the botrun flow lang API (LLM can get this from system prompt)
-        user_id: REQUIRED - User ID for file upload (LLM can get this from system prompt)
-
-    Returns:
-        str: A public URL that can be used to access the file for 7 days
-
-    Raises:
-        FileNotFoundError: If the specified file does not exist
-    """
-    logger.info(f"generate_tmp_public_url file_path: {file_path}")
-
-    if not os.path.exists(file_path):
-        return f"File not found: {file_path}"
-        # raise FileNotFoundError(f"File not found: {file_path}")
-
-    return upload_and_get_tmp_public_url(file_path, botrun_flow_lang_url, user_id)
-
-
-@mcp.tool()
-async def create_html_page(
-    html_content: str,
-    title: str,
-    botrun_flow_lang_url: str,
-    user_id: str,
-) -> str:
-    """
-    Create a custom HTML page and return its URL.
-
-    This tool supports complete HTML documents, including JavaScript and CSS, which can be used to create
-    complex interactive pages.
-
-    Prioritize using the following frameworks for writing HTML:
-    - DataTables for tables (include in header: <link rel="stylesheet" href="https://cdn.datatables.net/2.3.3/css/dataTables.dataTables.css" /> and <script src="https://cdn.datatables.net/2.3.3/js/dataTables.js"></script>)
-    - Alpine.js for interactivity (include in header: <script defer src="https://cdn.jsdelivr.net/npm/alpinejs@3.x.x/dist/cdn.min.js"></script>)
-    - Tailwind CSS for styling (include in header: <script src="https://cdn.jsdelivr.net/npm/@tailwindcss/browser@4"></script>)
-    - daisyUI for UI components (include in header: <link href="https://cdn.jsdelivr.net/npm/daisyui@5" rel="stylesheet" type="text/css" />)
-    - Chart.js for charts (include in header: <script src="https://cdn.jsdelivr.net/npm/chart.js"></script>)
-    - Animate.css for animations (include in header: <link rel="stylesheet" href="https://cdnjs.cloudflare.com/ajax/libs/animate.css/4.1.1/animate.min.css" />)
-
-    Input Options:
-    You can pass either:
-    1. A complete HTML document with doctype, html, head, and body tags
-    2. An HTML fragment that will be automatically wrapped in a basic HTML structure
-
-    Args:
-        html_content: Complete HTML document or fragment. Can include JavaScript, CSS, and other elements.
-        title: Optional title for the HTML page
-        botrun_flow_lang_url: REQUIRED - URL for the botrun flow lang API (LLM can get this from system prompt)
-        user_id: REQUIRED - User ID for file upload (LLM can get this from system prompt)
-
-    Returns:
-        str: URL for the HTML page. This URL should be provided to the user,
-             as they will need to access it to view the content in their web browser.
-    """
-    try:
-        logger.info(f"create_html_page html_content: {html_content} title: {title}")
-
-        html_url = await generate_html_file(
-            html_content, botrun_flow_lang_url, user_id, title
-        )
-
-        logger.info(f"create_html_page generated============> {html_url}")
-        return html_url
-
-    except Exception as e:
-        logger.error(f"create_html_page error: {e}", error=str(e), exc_info=True)
-        return f"Error creating HTML page URL: {str(e)}"
-
-
-@mcp.tool()
-async def create_plotly_chart(
-    figure_data: str | dict,
-    title: str,
-    botrun_flow_lang_url: str,
-    user_id: str,
-) -> str:
-    """
-    Create an interactive Plotly visualization and return its URL.
-    This URL should be provided to the user,
-    as they will need to access it to view the interactive chart in their web browser.
-
-    Scenarios for using create_plotly_chart:
-    - Need to create data visualizations and charts
-    - Need to show data trends (line charts)
-    - Need to compare values (bar charts, pie charts)
-    - Need to show distributions (scatter plots, heat maps)
-    - Need to display time series data (timeline charts)
-    - Need to show geographic information (maps)
-    - Need to perform multidimensional data analysis (3D charts, bubble charts)
-    - Need to show statistical distributions (box plots)
-    - Need to show cumulative trends (area charts)
-    - Need interactive data exploration
-
-    Integration with Other Tools:
-    This function can be used in conjunction with chat_with_imgs and chat_with_pdf when they return data
-    suitable for visualization. When those tools detect a need for visualization, they will return a JSON string
-    with a "__plotly_data__" key, which can be directly passed to this function.
-
-    Example workflow:
-    1. User asks to analyze and visualize data from images/PDFs
-    2. chat_with_imgs or chat_with_pdf returns JSON string with "__plotly_data__" key
-    3. Pass that string to this function to get an interactive visualization URL
-
-    Supported Chart Types:
-    - Line charts: For showing trends and time series data
-    - Bar charts: For comparing values across categories
-    - Pie charts: For showing proportions of a whole
-    - Scatter plots: For showing relationships between variables
-    - Heat maps: For showing patterns in matrix data
-    - Box plots: For showing statistical distributions
-    - Geographic maps: For showing spatial data
-    - 3D plots: For showing three-dimensional data
-    - Bubble charts: For showing three variables in 2D
-    - Area charts: For showing cumulative totals over time
-
-    The figure_data can be either:
-    1. A JSON string containing plotly figure specifications with 'data' and 'layout'
-    2. A dictionary object with plotly figure specifications
-
-    Example JSON string:
-    '{"data": [{"type": "scatter", "x": [1, 2, 3, 4], "y": [10, 15, 13, 17]}], "layout": {"title": "My Plot"}}'
-
-    Example dictionary:
-    {
-        'data': [{
-            'type': 'scatter',
-            'x': [1, 2, 3, 4],
-            'y': [10, 15, 13, 17]
-        }],
-        'layout': {
-            'title': 'My Plot'
-        }
-    }
-
-    Args:
-        figure_data: JSON string OR dictionary containing plotly figure specifications or output from chat_with_imgs/chat_with_pdf.
-                    String inputs will be parsed using json.loads(), dictionary inputs will be used directly.
-        title: Optional title for the plot.
-        botrun_flow_lang_url: REQUIRED - URL for the botrun flow lang API (LLM can get this from system prompt)
-        user_id: REQUIRED - User ID for file upload (LLM can get this from system prompt)
-
-    Returns:
-        str: URL for the interactive HTML visualization. This URL should be provided to the user,
-             as they will need to access it to view the interactive chart in their web browser.
-    """
-    try:
-        logger.info(f"create_plotly_chart figure_data: {figure_data} title: {title}")
-
-        # Handle both string and dict inputs
-        if isinstance(figure_data, str):
-            figure_dict = json.loads(figure_data)
-        else:
-            figure_dict = figure_data
-
-        # If the input is from chat_with_imgs or chat_with_pdf, extract the plotly data
-        if "__plotly_data__" in figure_dict:
-            figure_dict = figure_dict["__plotly_data__"]
-
-        html_url = await generate_plotly_files(
-            figure_dict,
-            botrun_flow_lang_url,
-            user_id,
-            title,
-        )
-
-        logger.info(f"create_plotly_chart generated============> {html_url}")
-        return html_url
-
-    except json.JSONDecodeError as e:
-        logger.error(
-            f"create_plotly_chart JSON parsing error: {e}", error=str(e), exc_info=True
-        )
-        return f"Error parsing JSON figure_data: {str(e)}. Please ensure figure_data is a valid JSON string or dictionary."
-    except KeyError as e:
-        logger.error(
-            f"create_plotly_chart missing key error: {e}", error=str(e), exc_info=True
-        )
-        return f"Error: Missing required key in figure data: {str(e)}. Please ensure figure_data contains 'data' and 'layout' keys."
-    except Exception as e:
-        logger.error(f"create_plotly_chart error: {e}", error=str(e), exc_info=True)
-        return f"Error creating visualization URL: {str(e)}"
-
-
-@mcp.tool()
-async def create_mermaid_diagram(
-    mermaid_data: str,
-    title: str,
-    botrun_flow_lang_url: str,
-    user_id: str,
-) -> str:
-    """
-    Create an interactive Mermaid diagram visualization and return its URL.
-    This URL should be provided to the user,
-    as they will need to access it to view the interactive diagram in their web browser.
-
-    Scenarios for using create_mermaid_diagram:
-    - Need to visualize flowcharts, architecture diagrams, or relationship diagrams
-    - Need to show system architecture (flowchart)
-    - Need to explain operational processes (flowchart)
-    - Need to show sequence interactions (sequence diagram)
-    - Need to show state transitions (state diagram)
-    - Need to show class relationships (class diagram)
-    - Need to show entity relationships (ER diagram)
-    - Need to show project timelines (gantt chart)
-    - Need to show user journeys (journey)
-    - Need to show requirement relationships (requirement diagram)
-    - Need to show resource allocation (pie chart)
-
-    Supported Diagram Types:
-    1. Flowcharts (graph TD/LR):
-       - System architectures
-       - Process flows
-       - Decision trees
-       - Data flows
-
-    2. Sequence Diagrams (sequenceDiagram):
-       - API interactions
-       - System communications
-       - User interactions
-       - Message flows
-
-    3. Class Diagrams (classDiagram):
-       - Software architecture
-       - Object relationships
-       - System components
-       - Code structure
-
-    4. State Diagrams (stateDiagram-v2):
-       - System states
-       - Workflow states
-       - Process states
-       - State transitions
-
-    5. Entity Relationship Diagrams (erDiagram):
-       - Database schemas
-       - Data relationships
-       - System entities
-       - Data models
-
-    6. User Journey Diagrams (journey):
-       - User experiences
-       - Customer flows
-       - Process steps
-       - Task sequences
-
-    7. Gantt Charts (gantt):
-       - Project timelines
-       - Task schedules
-       - Resource allocation
-       - Milestone tracking
-
-    8. Pie Charts (pie):
-       - Data distribution
-       - Resource allocation
-       - Market share
-       - Component breakdown
-
-    9. Requirement Diagrams (requirementDiagram):
-       - System requirements
-       - Dependencies
-       - Specifications
-       - Constraints
-
-    Example Mermaid syntax for a simple flowchart:
-    ```
-    graph TD
-        A[Start] --> B{Data Available?}
-        B -->|Yes| C[Process Data]
-        B -->|No| D[Get Data]
-        C --> E[End]
-        D --> B
-    ```
-
-    Args:
-        mermaid_data: String containing the Mermaid diagram definition
-        title: Optional title for the diagram
-        botrun_flow_lang_url: REQUIRED - URL for the botrun flow lang API (LLM can get this from system prompt)
-        user_id: REQUIRED - User ID for file upload (LLM can get this from system prompt)
-
-    Returns:
-        str: URL for the interactive HTML visualization. This URL should be provided to the user,
-             as they will need to access it to view the interactive diagram in their web browser.
-    """
-    try:
-        logger.info(
-            f"create_mermaid_diagram mermaid_data: {mermaid_data} title: {title}"
-        )
-
-        html_url = await generate_mermaid_files(
-            mermaid_data,
-            botrun_flow_lang_url,
-            user_id,
-            title,
-        )
-
-        logger.info(f"create_mermaid_diagram generated============> {html_url}")
-        return html_url
-
-    except Exception as e:
-        logger.error(f"create_mermaid_diagram error: {e}", error=str(e), exc_info=True)
-        return f"Error creating diagram URL: {str(e)}"
-
-
-@mcp.tool()
-async def current_date_time() -> str:
-    """
-    Get the current date and time in local timezone (Asia/Taipei).
-
-    Important: You MUST call this current_date_time function when:
-    1. User's query contains time-related words such as:
-       - today, now, current
-       - this week, next week
-       - this month, last month
-       - this year, last year, next year
-       - recent, lately
-       - future, upcoming
-       - past, previous
-    2. User asks about current events or latest information
-    3. User wants to know time-sensitive information
-    4. Queries involving relative time expressions
-
-    Examples of when to use current_date_time:
-    - "What's the weather today?"
-    - "This month's stock market performance"
-    - "Any recent news?"
-    - "Economic growth from last year until now"
-    - "Upcoming events for next week"
-    - "This month's sales data"
-
-    Args:
-        botrun_flow_lang_url: Optional URL for the botrun flow lang API (not used for this tool)
-        user_id: Optional user ID (not used for this tool)
-
-    Returns:
-        str: Current date and time in format "YYYY-MM-DD HH:MM Asia/Taipei"
-    """
-    try:
-        logger.info("current_date_time called")
-
-        local_tz = pytz.timezone("Asia/Taipei")
-        local_time = datetime.now(local_tz)
-        result = local_time.strftime("%Y-%m-%d %H:%M %Z")
-
-        logger.info(f"current_date_time============> {result}")
-        return result
-
-    except Exception as e:
-        logger.error(f"current_date_time error: {e}", error=str(e), exc_info=True)
-        return f"Error: {e}"
-
-
-def format_dates(dt):
-    """
-    Format datetime for both Western and Taiwan formats
-    Western format: yyyy-mm-dd hh:mm:ss
-    Taiwan format: (yyyy-1911)-mm-dd hh:mm:ss
-    """
-    western_date = dt.strftime("%Y-%m-%d %H:%M:%S")
-    taiwan_year = dt.year - 1911
-    taiwan_date = f"{taiwan_year}-{dt.strftime('%m-%d %H:%M:%S')}"
-
-    return {"western_date": western_date, "taiwan_date": taiwan_date}
-
-
-@mcp.tool()
-async def web_search(
-    user_input: str,
-    return_images: bool = False,
-) -> dict:
-    """
-    Search the web for up-to-date information using Perplexity.
-    This tool provides detailed answers with citations.
-
-    Unless the user insists on multiple-round searches, this tool can search for multiple conditions at once, for example:
-    - Good: web_search("Search for today's sports, financial, and political news")
-    - Unnecessary: Making separate searches for sports, financial, and political news
-
-    Time/Date Information Requirements:
-    1. MUST preserve any specific dates or time periods mentioned in the user's query
-    2. Include both the current time and any specific time references from the query
-
-    Examples:
-    - Basic query:
-      User asks: "Population of Japan"
-      web_search("Population of Japan")
-      Returns: {
-          "content": "According to the latest statistics, Japan's population is about 125 million...",
-          "citations": [
-              {"title": "Statistics Bureau of Japan", "url": "https://www.stat.go.jp/..."},
-              {"title": "World Bank Data", "url": "https://data.worldbank.org/..."}
-          ]
-      }
-
-    - Query with specific date:
-      User asks: "Look up news from January 15, 2023"
-      web_search("Look up news from January 15, 2023")
-      Returns: {
-          "content": "News from January 15, 2023 includes...",
-          "citations": [
-              {"title": "BBC News", "url": "https://www.bbc.com/..."},
-              {"title": "Reuters", "url": "https://www.reuters.com/..."}
-          ]
-      }
-
-    - Location-specific query:
-      User asks: "Weather in Paris today"
-      web_search("Weather in Paris today")
-      Returns: {
-          "content": "Today's weather in Paris shows...",
-          "citations": [
-              {"title": "Weather Service", "url": "https://www.weather.com/..."},
-              {"title": "Meteorological Office", "url": "https://www.metoffice.gov.uk/..."}
-          ]
-      }
-
-    Args:
-        user_input: The search query or question you want to find information about.
-                   MUST include any specific time periods or dates from the original query.
-                   Examples of time formats to preserve:
-                   - Specific dates: "2025/1/1", "2023-12-31", "January 15, 2023"
-                   - Years: "2023"
-                   - Quarters/Months: "Q1", "January", "First quarter"
-                   - Time periods: "past three years", "next five years"
-                   - Relative time: "yesterday", "next week", "last month"
-        return_images: Whether to include images in search results. Set to True when you need to search for and return images along with text content.
-        botrun_flow_lang_url: Optional URL for the botrun flow lang API (not used for this tool)
-        user_id: Optional user ID (not used for this tool)
-
-    Returns:
-        dict: A dictionary containing:
-              - content (str): The detailed answer based on web search results
-              - citations (list): A list of URLs, citations are important to provide to the user
-              - images (list): A list of image URLs (only when return_images is True)
-    """
-    try:
-        logger.info(f"web_search user_input: {user_input}")
-
-        now = datetime.now()
-        dates = format_dates(now)
-        western_date = dates["western_date"]
-        taiwan_date = dates["taiwan_date"]
-
-        logger.info(f"western_date: {western_date} taiwan_date: {taiwan_date}")
-
-        # Format input with current time context (English only)
-        final_input = (
-            f"The current date: {western_date}\nThe user's question is: {user_input}"
-        )
-
-        # Process search using async generator
-        search_result = {
-            "content": "",
-            "citations": [],
-        }
-
-        async for event in respond_with_perplexity_search(
-            final_input,
-            user_prompt_prefix="",
-            messages_for_llm=[],
-            domain_filter=[],
-            stream=False,
-            structured_output=True,
-            return_images=return_images,
-        ):
-            if event and isinstance(event.chunk, str):
-                search_result = json.loads(event.chunk)
-
-        logger.info(
-            f"web_search completed============> {len(search_result.get('content', ''))}"
-        )
-        return (
-            search_result
-            if search_result
-            else {"content": "No results found.", "citations": []}
-        )
-
-    except Exception as e:
-        logger.error(f"web_search error: {e}", error=str(e), exc_info=True)
-        return {"content": f"Error during web search: {str(e)}", "citations": []}
+import os
+import asyncio
+import json
+import pytz
+from datetime import datetime
+from typing import Dict, Any, Optional
+
+# Import for web_search
+from botrun_flow_lang.langgraph_agents.agents.util.perplexity_search import (
+    respond_with_perplexity_search,
+)
+
+from mcp.server.fastmcp import FastMCP
+from langchain_core.runnables import RunnableConfig
+
+# Import necessary dependencies
+from botrun_flow_lang.models.nodes.utils import scrape_single_url
+from botrun_flow_lang.langgraph_agents.agents.util.pdf_analyzer import (
+    analyze_pdf_async,
+)
+from botrun_flow_lang.langgraph_agents.agents.util.img_util import analyze_imgs
+from botrun_flow_lang.langgraph_agents.agents.util.local_files import (
+    upload_and_get_tmp_public_url,
+)
+from botrun_flow_lang.langgraph_agents.agents.util.html_util import generate_html_file
+from botrun_flow_lang.langgraph_agents.agents.util.plotly_util import (
+    generate_plotly_files,
+)
+from botrun_flow_lang.langgraph_agents.agents.util.mermaid_util import (
+    generate_mermaid_files,
+)
+from botrun_flow_lang.utils.clients.rate_limit_client import RateLimitClient
+from botrun_flow_lang.utils.botrun_logger import get_default_botrun_logger
+
+# Import for generate_image
+from langchain_community.utilities.dalle_image_generator import DallEAPIWrapper
+from langchain_community.callbacks import get_openai_callback
+
+# Initialize MCP server
+mcp = FastMCP(name="BotrunFlowLangDefaultMCP", stateless_http=True)
+
+# Initialize logger - reuse the default instance
+logger = get_default_botrun_logger()
+
+
+# Exception class for rate limit errors
+class BotrunRateLimitException(Exception):
+    """Exception that should be displayed directly to the user."""
+
+    def __init__(self, message):
+        self.message = f"[Please tell user error] {message}"
+        super().__init__(self.message)
+
+
+@mcp.tool()
+async def scrape(url: str) -> dict:
+    """
+    Scrape a web page to extract its content.
+
+    Args:
+        url: The URL to scrape
+
+    Returns:
+        dict: A dictionary containing the scraped content or error message
+    """
+    try:
+        logger.info(f"scrape {url}")
+        return await scrape_single_url(url)
+    except Exception as e:
+        logger.error(f"scrape {url} error: {e}", error=str(e), exc_info=True)
+        return {"url": url, "status": "error", "error": str(e)}
+
+
+@mcp.tool()
+async def chat_with_pdf(
+    pdf_url: str, user_input: str, botrun_flow_lang_url: str, user_id: str
+) -> str:
+    """
+    Analyze a PDF file and answer questions about its content.
+
+    Supports intelligent processing based on file size:
+    - Small files (< 5MB): Direct multimodal analysis
+    - Large files (>= 5MB): Compress -> Split -> Parallel multimodal Q&A -> Merge results
+
+    Args:
+        pdf_url: The URL to the PDF file (can be generated using generate_tmp_public_url for local files)
+        user_input: The user's question or instruction about the PDF content
+        botrun_flow_lang_url: REQUIRED - URL for the botrun flow lang API (LLM can get this from system prompt)
+        user_id: REQUIRED - User ID for file upload (LLM can get this from system prompt)
+
+    Returns:
+        str: Analysis result or Plotly-compatible data structure if visualization is needed
+    """
+    logger.info(f"chat_with_pdf pdf_url: {pdf_url} user_input: {user_input}")
+
+    # Convert local file path to public URL if needed
+    if not pdf_url.startswith("http"):
+        pdf_url = upload_and_get_tmp_public_url(pdf_url, botrun_flow_lang_url, user_id)
+
+    return await analyze_pdf_async(pdf_url, user_input)
+
+
+@mcp.tool()
+async def chat_with_imgs(
+    img_urls: list[str],
+    user_input: str,
+    botrun_flow_lang_url: str,
+    user_id: str,
+) -> str:
+    """
+    Analyze multiple images and answer questions about their content.
+
+    Args:
+        img_urls: List of URLs to the image files (can be generated using generate_tmp_public_url for local files)
+        user_input: Question or instruction about the image content(s)
+        botrun_flow_lang_url: REQUIRED - URL for the botrun flow lang API (LLM can get this from system prompt)
+        user_id: REQUIRED - User ID for file upload (LLM can get this from system prompt)
+
+    Returns:
+        str: Analysis result or Plotly-compatible data structure if visualization is needed
+    """
+    logger.info(f"chat_with_imgs img_urls: {img_urls} user_input: {user_input}")
+
+    # Convert local file paths to public URLs if needed
+    new_img_urls = []
+    for img_url in img_urls:
+        if not img_url.startswith("http"):
+            img_url = upload_and_get_tmp_public_url(
+                img_url, botrun_flow_lang_url, user_id
+            )
+        new_img_urls.append(img_url)
+
+    return analyze_imgs(new_img_urls, user_input)
+
+
+@mcp.tool()
+async def generate_image(
+    user_input: str, user_id: str = "", botrun_flow_lang_url: str = ""
+) -> str:
+    """
+    Generate high-quality images using DALL-E 3 and store permanently in GCS.
+
+    When using generate_image tool, you must include the image URL in your response.
+    You MUST respond using this format (from @begin img to @end, including the image URL):
+    @begin img("{image_url}") @end
+
+    Capabilities:
+    - Creates photorealistic images and art
+    - Handles complex scenes and compositions
+    - Maintains consistent styles
+    - Follows detailed prompts with high accuracy
+    - Supports various artistic styles and mediums
+
+    Best practices for prompts:
+    - Be specific about style, mood, lighting, and composition
+    - Include details about perspective and setting
+    - Specify artistic medium if desired (e.g., "oil painting", "digital art")
+    - Mention color schemes or specific colors
+    - Describe the atmosphere or emotion you want to convey
+
+    Limitations:
+    - Cannot generate images of public figures or celebrities
+    - Avoids harmful, violent, or adult content
+    - May have inconsistencies with hands, faces, or text
+    - Cannot generate exact copies of existing artworks or brands
+    - Limited to single image generation per request
+    - Subject to daily usage limits
+
+    Args:
+        user_input: Detailed description of the image you want to generate.
+                   Be specific about style, content, and composition.
+        user_id: REQUIRED - User ID for rate limit and file storage (LLM can get this from system prompt)
+        botrun_flow_lang_url: REQUIRED - URL for the botrun flow lang API (LLM can get this from system prompt)
+
+    Returns:
+        str: Permanent URL to the generated image stored in GCS, or error message if generation fails
+    """
+    try:
+        logger.info(f"generate_image user_input: {user_input}")
+
+        # 驗證必要參數
+        if not user_id:
+            logger.error("User ID not available")
+            return "User ID not available"
+        if not botrun_flow_lang_url:
+            logger.error("botrun_flow_lang_url not available")
+            return "botrun_flow_lang_url not available"
+
+        # Check rate limit before generating image
+        rate_limit_client = RateLimitClient()
+        rate_limit_info = await rate_limit_client.get_rate_limit(user_id)
+
+        # Check if user can generate an image
+        drawing_info = rate_limit_info.get("drawing", {})
+        can_use = drawing_info.get("can_use", False)
+
+        if not can_use:
+            daily_limit = drawing_info.get("daily_limit", 0)
+            current_usage = drawing_info.get("current_usage", 0)
+            logger.error(
+                f"User {user_id} has reached daily limit of {daily_limit} image generations. "
+                f"Current usage: {current_usage}. Please try again tomorrow."
+            )
+            return f"[Please tell user error] You have reached your daily limit of {daily_limit} image generations. " \
+                f"Current usage: {current_usage}. Please try again tomorrow."
+            # raise BotrunRateLimitException(
+            #     f"You have reached your daily limit of {daily_limit} image generations. "
+            #     f"Current usage: {current_usage}. Please try again tomorrow."
+            # )
+
+        # 2. 使用 DALL-E 生成圖片
+        dalle_wrapper = DallEAPIWrapper(
+            api_key=os.getenv("OPENAI_API_KEY"), model="dall-e-3"
+        )
+
+        # Generate image with token usage tracking
+        with get_openai_callback() as cb:
+            temp_image_url = dalle_wrapper.run(user_input)
+            logger.info(
+                f"DALL-E generated temporary URL: {temp_image_url}, "
+                f"prompt tokens: {cb.prompt_tokens}, "
+                f"completion tokens: {cb.completion_tokens}"
+            )
+
+        # 3. 下載並上傳到 GCS，取得永久 URL
+        from botrun_flow_lang.langgraph_agents.agents.util.local_files import (
+            upload_image_and_get_public_url,
+        )
+
+        try:
+            permanent_url = await upload_image_and_get_public_url(
+                temp_image_url, botrun_flow_lang_url, user_id
+            )
+            logger.info(f"Image uploaded to GCS: {permanent_url}")
+
+            # 4. 更新使用計數
+            await rate_limit_client.update_drawing_usage(user_id)
+
+            return permanent_url
+        except Exception as upload_error:
+            logger.error(
+                f"Failed to upload to GCS, returning temporary URL: {upload_error}"
+            )
+            # Fallback: 回傳臨時 URL
+            await rate_limit_client.update_drawing_usage(user_id)
+            return temp_image_url
+
+    except Exception as e:
+        logger.error(f"generate_image error: {e}", error=str(e), exc_info=True)
+
+        # Check if this is a user-visible exception
+        if str(e).startswith("[Please tell user error]"):
+            return str(e)  # Return the error message as is
+        return f"Error: {e}"
+
+
+@mcp.tool()
+async def generate_tmp_public_url(
+    file_path: str, botrun_flow_lang_url: str, user_id: str
+) -> str:
+    """
+    Create a temporary public URL for a local file.
+
+    Args:
+        file_path: The path to the local file you want to make publicly accessible
+        botrun_flow_lang_url: REQUIRED - URL for the botrun flow lang API (LLM can get this from system prompt)
+        user_id: REQUIRED - User ID for file upload (LLM can get this from system prompt)
+
+    Returns:
+        str: A public URL that can be used to access the file for 7 days
+
+    Raises:
+        FileNotFoundError: If the specified file does not exist
+    """
+    logger.info(f"generate_tmp_public_url file_path: {file_path}")
+
+    if not os.path.exists(file_path):
+        return f"File not found: {file_path}"
+        # raise FileNotFoundError(f"File not found: {file_path}")
+
+    return upload_and_get_tmp_public_url(file_path, botrun_flow_lang_url, user_id)
+
+
+@mcp.tool()
+async def create_html_page(
+    html_content: str,
+    title: str,
+    botrun_flow_lang_url: str,
+    user_id: str,
+) -> str:
+    """
+    Create a custom HTML page and return its URL.
+
+    This tool supports complete HTML documents, including JavaScript and CSS, which can be used to create
+    complex interactive pages.
+
+    Prioritize using the following frameworks for writing HTML:
+    - DataTables for tables (include in header: <link rel="stylesheet" href="https://cdn.datatables.net/2.3.3/css/dataTables.dataTables.css" /> and <script src="https://cdn.datatables.net/2.3.3/js/dataTables.js"></script>)
+    - Alpine.js for interactivity (include in header: <script defer src="https://cdn.jsdelivr.net/npm/alpinejs@3.x.x/dist/cdn.min.js"></script>)
+    - Tailwind CSS for styling (include in header: <script src="https://cdn.jsdelivr.net/npm/@tailwindcss/browser@4"></script>)
+    - daisyUI for UI components (include in header: <link href="https://cdn.jsdelivr.net/npm/daisyui@5" rel="stylesheet" type="text/css" />)
+    - Chart.js for charts (include in header: <script src="https://cdn.jsdelivr.net/npm/chart.js"></script>)
+    - Animate.css for animations (include in header: <link rel="stylesheet" href="https://cdnjs.cloudflare.com/ajax/libs/animate.css/4.1.1/animate.min.css" />)
+
+    Input Options:
+    You can pass either:
+    1. A complete HTML document with doctype, html, head, and body tags
+    2. An HTML fragment that will be automatically wrapped in a basic HTML structure
+
+    Args:
+        html_content: Complete HTML document or fragment. Can include JavaScript, CSS, and other elements.
+        title: Optional title for the HTML page
+        botrun_flow_lang_url: REQUIRED - URL for the botrun flow lang API (LLM can get this from system prompt)
+        user_id: REQUIRED - User ID for file upload (LLM can get this from system prompt)
+
+    Returns:
+        str: URL for the HTML page. This URL should be provided to the user,
+             as they will need to access it to view the content in their web browser.
+    """
+    try:
+        logger.info(f"create_html_page html_content: {html_content} title: {title}")
+
+        html_url = await generate_html_file(
+            html_content, botrun_flow_lang_url, user_id, title
+        )
+
+        logger.info(f"create_html_page generated============> {html_url}")
+        return html_url
+
+    except Exception as e:
+        logger.error(f"create_html_page error: {e}", error=str(e), exc_info=True)
+        return f"Error creating HTML page URL: {str(e)}"
+
+
+@mcp.tool()
+async def create_plotly_chart(
+    figure_data: str | dict,
+    title: str,
+    botrun_flow_lang_url: str,
+    user_id: str,
+) -> str:
+    """
+    Create an interactive Plotly visualization and return its URL.
+    This URL should be provided to the user,
+    as they will need to access it to view the interactive chart in their web browser.
+
+    Scenarios for using create_plotly_chart:
+    - Need to create data visualizations and charts
+    - Need to show data trends (line charts)
+    - Need to compare values (bar charts, pie charts)
+    - Need to show distributions (scatter plots, heat maps)
+    - Need to display time series data (timeline charts)
+    - Need to show geographic information (maps)
+    - Need to perform multidimensional data analysis (3D charts, bubble charts)
+    - Need to show statistical distributions (box plots)
+    - Need to show cumulative trends (area charts)
+    - Need interactive data exploration
+
+    Integration with Other Tools:
+    This function can be used in conjunction with chat_with_imgs and chat_with_pdf when they return data
+    suitable for visualization. When those tools detect a need for visualization, they will return a JSON string
+    with a "__plotly_data__" key, which can be directly passed to this function.
+
+    Example workflow:
+    1. User asks to analyze and visualize data from images/PDFs
+    2. chat_with_imgs or chat_with_pdf returns JSON string with "__plotly_data__" key
+    3. Pass that string to this function to get an interactive visualization URL
+
+    Supported Chart Types:
+    - Line charts: For showing trends and time series data
+    - Bar charts: For comparing values across categories
+    - Pie charts: For showing proportions of a whole
+    - Scatter plots: For showing relationships between variables
+    - Heat maps: For showing patterns in matrix data
+    - Box plots: For showing statistical distributions
+    - Geographic maps: For showing spatial data
+    - 3D plots: For showing three-dimensional data
+    - Bubble charts: For showing three variables in 2D
+    - Area charts: For showing cumulative totals over time
+
+    The figure_data can be either:
+    1. A JSON string containing plotly figure specifications with 'data' and 'layout'
+    2. A dictionary object with plotly figure specifications
+
+    Example JSON string:
+    '{"data": [{"type": "scatter", "x": [1, 2, 3, 4], "y": [10, 15, 13, 17]}], "layout": {"title": "My Plot"}}'
+
+    Example dictionary:
+    {
+        'data': [{
+            'type': 'scatter',
+            'x': [1, 2, 3, 4],
+            'y': [10, 15, 13, 17]
+        }],
+        'layout': {
+            'title': 'My Plot'
+        }
+    }
+
+    Args:
+        figure_data: JSON string OR dictionary containing plotly figure specifications or output from chat_with_imgs/chat_with_pdf.
+                    String inputs will be parsed using json.loads(), dictionary inputs will be used directly.
+        title: Optional title for the plot.
+        botrun_flow_lang_url: REQUIRED - URL for the botrun flow lang API (LLM can get this from system prompt)
+        user_id: REQUIRED - User ID for file upload (LLM can get this from system prompt)
+
+    Returns:
+        str: URL for the interactive HTML visualization. This URL should be provided to the user,
+             as they will need to access it to view the interactive chart in their web browser.
+    """
+    try:
+        logger.info(f"create_plotly_chart figure_data: {figure_data} title: {title}")
+
+        # Handle both string and dict inputs
+        if isinstance(figure_data, str):
+            figure_dict = json.loads(figure_data)
+        else:
+            figure_dict = figure_data
+
+        # If the input is from chat_with_imgs or chat_with_pdf, extract the plotly data
+        if "__plotly_data__" in figure_dict:
+            figure_dict = figure_dict["__plotly_data__"]
+
+        html_url = await generate_plotly_files(
+            figure_dict,
+            botrun_flow_lang_url,
+            user_id,
+            title,
+        )
+
+        logger.info(f"create_plotly_chart generated============> {html_url}")
+        return html_url
+
+    except json.JSONDecodeError as e:
+        logger.error(
+            f"create_plotly_chart JSON parsing error: {e}", error=str(e), exc_info=True
+        )
+        return f"Error parsing JSON figure_data: {str(e)}. Please ensure figure_data is a valid JSON string or dictionary."
+    except KeyError as e:
+        logger.error(
+            f"create_plotly_chart missing key error: {e}", error=str(e), exc_info=True
+        )
+        return f"Error: Missing required key in figure data: {str(e)}. Please ensure figure_data contains 'data' and 'layout' keys."
+    except Exception as e:
+        logger.error(f"create_plotly_chart error: {e}", error=str(e), exc_info=True)
+        return f"Error creating visualization URL: {str(e)}"
+
+
+@mcp.tool()
+async def create_mermaid_diagram(
+    mermaid_data: str,
+    title: str,
+    botrun_flow_lang_url: str,
+    user_id: str,
+) -> str:
+    """
+    Create an interactive Mermaid diagram visualization and return its URL.
+    This URL should be provided to the user,
+    as they will need to access it to view the interactive diagram in their web browser.
+
+    Scenarios for using create_mermaid_diagram:
+    - Need to visualize flowcharts, architecture diagrams, or relationship diagrams
+    - Need to show system architecture (flowchart)
+    - Need to explain operational processes (flowchart)
+    - Need to show sequence interactions (sequence diagram)
+    - Need to show state transitions (state diagram)
+    - Need to show class relationships (class diagram)
+    - Need to show entity relationships (ER diagram)
+    - Need to show project timelines (gantt chart)
+    - Need to show user journeys (journey)
+    - Need to show requirement relationships (requirement diagram)
+    - Need to show resource allocation (pie chart)
+
+    Supported Diagram Types:
+    1. Flowcharts (graph TD/LR):
+       - System architectures
+       - Process flows
+       - Decision trees
+       - Data flows
+
+    2. Sequence Diagrams (sequenceDiagram):
+       - API interactions
+       - System communications
+       - User interactions
+       - Message flows
+
+    3. Class Diagrams (classDiagram):
+       - Software architecture
+       - Object relationships
+       - System components
+       - Code structure
+
+    4. State Diagrams (stateDiagram-v2):
+       - System states
+       - Workflow states
+       - Process states
+       - State transitions
+
+    5. Entity Relationship Diagrams (erDiagram):
+       - Database schemas
+       - Data relationships
+       - System entities
+       - Data models
+
+    6. User Journey Diagrams (journey):
+       - User experiences
+       - Customer flows
+       - Process steps
+       - Task sequences
+
+    7. Gantt Charts (gantt):
+       - Project timelines
+       - Task schedules
+       - Resource allocation
+       - Milestone tracking
+
+    8. Pie Charts (pie):
+       - Data distribution
+       - Resource allocation
+       - Market share
+       - Component breakdown
+
+    9. Requirement Diagrams (requirementDiagram):
+       - System requirements
+       - Dependencies
+       - Specifications
+       - Constraints
+
+    Example Mermaid syntax for a simple flowchart:
+    ```
+    graph TD
+        A[Start] --> B{Data Available?}
+        B -->|Yes| C[Process Data]
+        B -->|No| D[Get Data]
+        C --> E[End]
+        D --> B
+    ```
+
+    Args:
+        mermaid_data: String containing the Mermaid diagram definition
+        title: Optional title for the diagram
+        botrun_flow_lang_url: REQUIRED - URL for the botrun flow lang API (LLM can get this from system prompt)
+        user_id: REQUIRED - User ID for file upload (LLM can get this from system prompt)
+
+    Returns:
+        str: URL for the interactive HTML visualization. This URL should be provided to the user,
+             as they will need to access it to view the interactive diagram in their web browser.
+    """
+    try:
+        logger.info(
+            f"create_mermaid_diagram mermaid_data: {mermaid_data} title: {title}"
+        )
+
+        html_url = await generate_mermaid_files(
+            mermaid_data,
+            botrun_flow_lang_url,
+            user_id,
+            title,
+        )
+
+        logger.info(f"create_mermaid_diagram generated============> {html_url}")
+        return html_url
+
+    except Exception as e:
+        logger.error(f"create_mermaid_diagram error: {e}", error=str(e), exc_info=True)
+        return f"Error creating diagram URL: {str(e)}"
+
+
+@mcp.tool()
+async def current_date_time() -> str:
+    """
+    Get the current date and time in local timezone (Asia/Taipei).
+
+    Important: You MUST call this current_date_time function when:
+    1. User's query contains time-related words such as:
+       - today, now, current
+       - this week, next week
+       - this month, last month
+       - this year, last year, next year
+       - recent, lately
+       - future, upcoming
+       - past, previous
+    2. User asks about current events or latest information
+    3. User wants to know time-sensitive information
+    4. Queries involving relative time expressions
+
+    Examples of when to use current_date_time:
+    - "What's the weather today?"
+    - "This month's stock market performance"
+    - "Any recent news?"
+    - "Economic growth from last year until now"
+    - "Upcoming events for next week"
+    - "This month's sales data"
+
+    Args:
+        botrun_flow_lang_url: Optional URL for the botrun flow lang API (not used for this tool)
+        user_id: Optional user ID (not used for this tool)
+
+    Returns:
+        str: Current date and time in format "YYYY-MM-DD HH:MM Asia/Taipei"
+    """
+    try:
+        logger.info("current_date_time called")
+
+        local_tz = pytz.timezone("Asia/Taipei")
+        local_time = datetime.now(local_tz)
+        result = local_time.strftime("%Y-%m-%d %H:%M %Z")
+
+        logger.info(f"current_date_time============> {result}")
+        return result
+
+    except Exception as e:
+        logger.error(f"current_date_time error: {e}", error=str(e), exc_info=True)
+        return f"Error: {e}"
+
+
+def format_dates(dt):
+    """
+    Format datetime for both Western and Taiwan formats
+    Western format: yyyy-mm-dd hh:mm:ss
+    Taiwan format: (yyyy-1911)-mm-dd hh:mm:ss
+    """
+    western_date = dt.strftime("%Y-%m-%d %H:%M:%S")
+    taiwan_year = dt.year - 1911
+    taiwan_date = f"{taiwan_year}-{dt.strftime('%m-%d %H:%M:%S')}"
+
+    return {"western_date": western_date, "taiwan_date": taiwan_date}
+
+
+@mcp.tool()
+async def web_search(
+    user_input: str,
+    return_images: bool = False,
+) -> dict:
+    """
+    Search the web for up-to-date information using Perplexity.
+    This tool provides detailed answers with citations.
+
+    Unless the user insists on multiple-round searches, this tool can search for multiple conditions at once, for example:
+    - Good: web_search("Search for today's sports, financial, and political news")
+    - Unnecessary: Making separate searches for sports, financial, and political news
+
+    Time/Date Information Requirements:
+    1. MUST preserve any specific dates or time periods mentioned in the user's query
+    2. Include both the current time and any specific time references from the query
+
+    Examples:
+    - Basic query:
+      User asks: "Population of Japan"
+      web_search("Population of Japan")
+      Returns: {
+          "content": "According to the latest statistics, Japan's population is about 125 million...",
+          "citations": [
+              {"title": "Statistics Bureau of Japan", "url": "https://www.stat.go.jp/..."},
+              {"title": "World Bank Data", "url": "https://data.worldbank.org/..."}
+          ]
+      }
+
+    - Query with specific date:
+      User asks: "Look up news from January 15, 2023"
+      web_search("Look up news from January 15, 2023")
+      Returns: {
+          "content": "News from January 15, 2023 includes...",
+          "citations": [
+              {"title": "BBC News", "url": "https://www.bbc.com/..."},
+              {"title": "Reuters", "url": "https://www.reuters.com/..."}
+          ]
+      }
+
+    - Location-specific query:
+      User asks: "Weather in Paris today"
+      web_search("Weather in Paris today")
+      Returns: {
+          "content": "Today's weather in Paris shows...",
+          "citations": [
+              {"title": "Weather Service", "url": "https://www.weather.com/..."},
+              {"title": "Meteorological Office", "url": "https://www.metoffice.gov.uk/..."}
+          ]
+      }
+
+    Args:
+        user_input: The search query or question you want to find information about.
+                   MUST include any specific time periods or dates from the original query.
+                   Examples of time formats to preserve:
+                   - Specific dates: "2025/1/1", "2023-12-31", "January 15, 2023"
+                   - Years: "2023"
+                   - Quarters/Months: "Q1", "January", "First quarter"
+                   - Time periods: "past three years", "next five years"
+                   - Relative time: "yesterday", "next week", "last month"
+        return_images: Whether to include images in search results. Set to True when you need to search for and return images along with text content.
+        botrun_flow_lang_url: Optional URL for the botrun flow lang API (not used for this tool)
+        user_id: Optional user ID (not used for this tool)
+
+    Returns:
+        dict: A dictionary containing:
+              - content (str): The detailed answer based on web search results
+              - citations (list): A list of URLs, citations are important to provide to the user
+              - images (list): A list of image URLs (only when return_images is True)
+    """
+    try:
+        logger.info(f"web_search user_input: {user_input}")
+
+        now = datetime.now()
+        dates = format_dates(now)
+        western_date = dates["western_date"]
+        taiwan_date = dates["taiwan_date"]
+
+        logger.info(f"western_date: {western_date} taiwan_date: {taiwan_date}")
+
+        # Format input with current time context (English only)
+        final_input = (
+            f"The current date: {western_date}\nThe user's question is: {user_input}"
+        )
+
+        # Process search using async generator
+        search_result = {
+            "content": "",
+            "citations": [],
+        }
+
+        async for event in respond_with_perplexity_search(
+            final_input,
+            user_prompt_prefix="",
+            messages_for_llm=[],
+            domain_filter=[],
+            stream=False,
+            structured_output=True,
+            return_images=return_images,
+        ):
+            if event and isinstance(event.chunk, str):
+                search_result = json.loads(event.chunk)
+
+        logger.info(
+            f"web_search completed============> {len(search_result.get('content', ''))}"
+        )
+        return (
+            search_result
+            if search_result
+            else {"content": "No results found.", "citations": []}
+        )
+
+    except Exception as e:
+        logger.error(f"web_search error: {e}", error=str(e), exc_info=True)
+        return {"content": f"Error during web search: {str(e)}", "citations": []}