Product Visual Content Package - Confirm & Copy Final Agent Code

Describe

Describe your task

Refine

Refine the plan

SubAgents

Review all agents

Deploy

Deploy your agent

Review Final Agent Copy

Expand

[
  {
    "taskID": 100,
    "semanticTitleOfTask": "Generate 10-second scrolling screenshot video of a specified website",
    "taskDescription": "This task obtains and records a short 10-second video of the provided website URL, scrolling from top to bottom to capture the entire visible content, and returns the link to the resulting MP4 file.",
    "inputDescription": "This task requires a single input: the URL of the website that should be visually captured. The system navigates to the page, automatically scrolls down while recording, and produces a 10-second video.",
    "inputRequired": [
      "variable2"
    ],
    "outputDescription": "The skill returns a URL pointing to an MP4 video containing the scrolling capture of the entire webpage.",
    "outputName": "website_video",
    "promptInstruction": ""
  },
  {
    "taskID": 190,
    "semanticTitleOfTask": "Text Rewriter - YouTube URL Constructor",
    "taskDescription": "This task uses a text rewriting approach to produce a properly formatted YouTube search URL for the provided product name. It transforms the input text by replacing spaces with plus signs (+) and returns the final URL that can be used to search for that product on YouTube.",
    "inputDescription": "The only input is the product name [variable1], which must be integrated into the YouTube search URL template. Any spaces in [variable1] should be replaced by plus signs (+).",
    "inputRequired": [
      "variable1"
    ],
    "outputDescription": "A plain text string containing the fully formed YouTube search URL needed for the next task.",
    "outputName": "youtube_search_url",
    "promptInstruction": "Please take the product name `variable1` and construct a complete YouTube search URL. The URL should follow this exact format: https://www.youtube.com/results?search_query=productNameHere. Replace any spaces in the product name with plus signs (+). Return only the complete URL, nothing else."
  },
  {
    "taskID": 191,
    "semanticTitleOfTask": "Website Screenshot Video Capture",
    "taskDescription": "This task uses a technical tool to create a short (approximately 10-second) scrolling screenshot video of the provided YouTube search URL. The generated output is an MP4 file URL showing the scrolling capture.",
    "inputDescription": "The input is the [youtube_search_url] generated by Task #190. The tool will use this URL and produce a video capture of the page's content as it scrolls from top to bottom.",
    "inputRequired": [
      "youtube_search_url"
    ],
    "outputDescription": "A URL to an MP4 video showing a scrolling capture of the YouTube search results page for the given product name.",
    "outputName": "youtube_screenshot_video",
    "promptInstruction": "No custom LLM prompt is required. This is a technical capture tool that processes the provided URL automatically."
  },
  {
    "taskID": 187,
    "semanticTitleOfTask": "Retrieve top 5 relevant images using Brave Search",
    "taskDescription": "This task retrieves up to 5 relevant image URLs from a Brave image search for the specified product name. The user must supply the product name as input, and the output is a list of up to 5 image URLs.",
    "inputDescription": "One input - the product name (variable1) to be used in the Brave image search query. The skill uses this input to find relevant images.",
    "inputRequired": [
      "variable1"
    ],
    "outputDescription": "An array of up to 5 URLs representing relevant images for the given product name.",
    "outputName": "image_urls",
    "promptInstruction": ""
  },
  {
    "taskID": 224,
    "semanticTitleOfTask": "Oracle News Article Query for product coverage",
    "taskDescription": "This task queries a specialized news oracle, returning a summary of at least three relevant and recent articles about the product. Each article entry must include the complete URL and a concise summary of the article’s key points.",
    "inputDescription": "One input: the product name or keyword. The Oracle uses this to locate relevant news coverage and compile the summary with URLs.",
    "inputRequired": [
      "variable1"
    ],
    "outputDescription": "A text response listing at least three article URLs with summaries about the product (oracle-news-summary).",
    "outputName": "oracle_news_summary",
    "promptInstruction": "What are the most recent and relevant news articles about PRODUCT_NAME? Please provide a comprehensive summary of each article and you MUST include the complete URLs. Return at least 3 articles, focusing on the most significant coverage. Format each entry with the URL followed by a brief summary of the key points."
  },
  {
    "taskID": 225,
    "semanticTitleOfTask": "News URL Extractor from Oracle summary",
    "taskDescription": "This task analyzes the Oracle's news summaries and extracts two specific URLs for the most relevant articles based on recency and direct relation to the product. It outputs only the two URLs, one per line.",
    "inputDescription": "One input: the multi-article summary from the Oracle, containing article URLs. The system must parse it to find exactly two relevant article URLs.",
    "inputRequired": [
      "oracle_news_summary"
    ],
    "outputDescription": "A short text listing exactly two URLs of the most relevant news articles (oracle-news-urls).",
    "outputName": "oracle_news_urls",
    "promptInstruction": "From the provided news article summaries, please extract exactly two URLs for the most relevant articles. Consider recency and relevance to PRODUCT_NAME in your selection. Return only the two URLs, one per line, without any additional commentary or formatting."
  },
  {
    "taskID": 226,
    "semanticTitleOfTask": "First News Article Video Capture",
    "taskDescription": "This task takes the first extracted news article URL and generates a short scrolling screenshot video as an MP4, capturing the article’s page content from top to bottom.",
    "inputDescription": "One input: the first URL from the extracted news article URLs. The tool uses this URL to navigate and record a short scrolling capture.",
    "inputRequired": [
      "oracle_news_urls"
    ],
    "outputDescription": "An MP4 video file URL that shows the scrolling page capture of the first news article (news-screenshot-video1).",
    "outputName": "news_screenshot_video1",
    "promptInstruction": ""
  },
  {
    "taskID": 227,
    "semanticTitleOfTask": "Second News Article Video Capture",
    "taskDescription": "This task takes the second extracted news article URL and generates a short scrolling screenshot video as an MP4, capturing the article’s page content from top to bottom.",
    "inputDescription": "One input: the second URL from the extracted news article URLs. The tool uses this URL to navigate and record a short scrolling capture.",
    "inputRequired": [
      "oracle_news_urls"
    ],
    "outputDescription": "An MP4 video file URL that shows the scrolling page capture of the second news article (news-screenshot-video2).",
    "outputName": "news_screenshot_video2",
    "promptInstruction": ""
  },
  {
    "taskID": 228,
    "semanticTitleOfTask": "Final JSON Compiler - news videos array",
    "taskDescription": "This task gathers the two news article scrolling video URLs (news-screenshot-video1, news-screenshot-video2) and compiles them into a single JSON object containing a list of the two video URLs.",
    "inputDescription": "Two inputs: the video URLs from the first and second news articles. They must be returned as part of a JSON object with a 'news_videos' array field.",
    "inputRequired": [
      "news_screenshot_video1",
      "news_screenshot_video2"
    ],
    "outputDescription": "A final text output in strict JSON format with a key 'news_videos' containing the two news article video URLs in an array (final-news-videos).",
    "outputName": "final_news_videos",
    "promptInstruction": "Please format the following video URLs into a clean JSON structure. The final output should be a JSON object with a single key 'news_videos' containing an array of exactly two video URLs. These URLs are for news article screenshot videos about PRODUCT_NAME. Return only the JSON structure, without any additional commentary."
  },
  {
    "taskID": 229,
    "semanticTitleOfTask": "LLM - Compile Final JSON Output of Media References",
    "taskDescription": "This task uses an LLM to gather and compile all media references (website video, YouTube video, images, and news article videos) into a standardized JSON structure. The final output consolidates these URLs into a single JSON object.",
    "inputDescription": "This task requires the following inputs retrieved from previous subagents: one website screenshot video URL, one YouTube search results video URL, up to five product image URLs, and two news article screenshot video URLs.",
    "inputRequired": [
      "website_video",
      "youtube_screenshot_video",
      "image_urls",
      "news_screenshot_video1",
      "news_screenshot_video2"
    ],
    "outputDescription": "Returns a single JSON object with the compiled URLs under the keys 'website-video', 'youtube-screenshot-video', 'image-urls', and 'final-news-videos'.",
    "outputName": "final_json_output",
    "promptInstruction": "Please compile the following media references into a JSON structure. Use the exact URLs provided without modification, preserving their order and text. Include these fields: 'website-video' for the main site video, 'youtube-screenshot-video' for the YouTube search results video, 'image-urls' for the product images, and 'final-news-videos' for the two news article videos. Return the output in this JSON format: { \"website-video\": \"...\", \"youtube-screenshot-video\": \"...\", \"image-urls\": [\"...\"], \"final-news-videos\": [\"...\", \"...\"] }."
  }
]

Happy? Now Copy-Paste To Proceed...

BACK TO SUBAGENTS

COPY TO CLIPBOARD