[task-title] - 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

Here is the updated and combined JSON array for all the subagents based on the provided guidelines:

[
  {
    "taskID": 169,
    "semanticTitleOfTask": "Create Dalle Image (Rectangle)",
    "taskDescription": "This skill creates a 1792x1024 square image with DalleE, based on a given text prompt. Dalle3 is a great AI image creator, especially for \"non photo realistic\" images. Once the image is created it is saved to the server as a .png file.",
    "inputDescription": "10-300 character text prompt",
    "inputRequired": [
      "VARIABLE1"
    ],
    "outputDescription": "internal and external link to the image",
    "outputName": "rectangular_image_url",
    "promptInstruction": "Create a 1792x1024 rectangular image with Dalle based on this text prompt: [VARIABLE1]"
  },
  {
    "taskID": 171,
    "semanticTitleOfTask": "Create Dalle Image (Transparent Square)",
    "taskDescription": "This skill creates a 1024x1024, square, transparent PNG file, based on a given text prompt. The image created should be of a singular object/scene with a clearly defined edge (so the background remover works particularly well). This skill is especially great at creating icons, vectors, cartoons, physical objects, etc. Once the image is created it is saved to the server as a .png file.",
    "inputDescription": "10-300 character text prompt",
    "inputRequired": [
      "VARIABLE2"
    ],
    "outputDescription": "internal and external link to the image",
    "outputName": "square_transparent_image_url",
    "promptInstruction": "Create a 1024x1024 square transparent PNG image with Dalle based on this text prompt: [VARIABLE2]. The image should be of a single object or scene with a clearly defined edge."
  },
  {
    "taskID": 217,
    "semanticTitleOfTask": "Select Most Relevant Template To Use In Next Prompt(s)",
    "taskDescription": "This skill will input a series of templates, and select the most relevant one based on a given description of the required task.",
    "inputDescription": "description of the task",
    "inputRequired": [
      "VARIABLE1",
      "VARIABLE2"  
    ],
    "outputDescription": "template (text, code.. some non-binary text template output of some kind)",
    "outputName": "selected_template",
    "promptInstruction": "Given the following task description: [VARIABLE1], select the most relevant template from these options: [VARIABLE2]"
  },
  {
    "taskID": 178,
    "semanticTitleOfTask": "Rewrite Text Based On Instructions",
    "taskDescription": "This skill uses a LLM (typically GPT4) to return some type of text, based on prompt instructions. This is a fundamental and important skill.",
    "inputDescription": "",
    "inputRequired": [
      "selected_template",
      "VARIABLE1"
    ], 
    "outputDescription": "Outputted text (or save EVERYTHING AS LINK and then import everything from link? May make it easier, and only adds a few seconds to each call? but just means like 100x the files being on server! hmmm)",
    "outputName": "rewritten_text",
    "promptInstruction": "Rewrite the following text based on these instructions: [VARIABLE1]\n\nOriginal text: [selected_template]"
  },
  {
    "taskID": 179,
    "semanticTitleOfTask": "Rewrite HTML Based On Instructions",
    "taskDescription": "This skill inputs some HTML (such as a template that has been pre-created) and uses a LLM (typically GPT4) to edit it, based on prompt instructions given. The HTML is then edited, as per the instructions, and returned.",
    "inputDescription": "HTML code AND the elements to be added into it AND instructions",
    "inputRequired": [
      "selected_template",
      "rewritten_text", 
      "VARIABLE1"
    ],
    "outputDescription": "HTML link (I think? or just use HTML itself? not sure? probably do it direct?)",
    "outputName": "edited_html",
    "promptInstruction": "Edit the following HTML code based on these instructions: [VARIABLE1]. Add these elements into the HTML where appropriate: [rewritten_text]\n\nOriginal HTML:\n[selected_template]"
  },
  {
    "taskID": 183,
    "semanticTitleOfTask": "Find Most Relevant YouTube Video For Keyword (Or Topic)",
    "taskDescription": "This skill inputs a keyword or topic and returns a list of the most relevant YouTube videos for that given task.",
    "inputDescription": "a single keyword",
    "inputRequired": [
      "VARIABLE1"  
    ],
    "outputDescription": "list of youtube videos ranked in order of relevance, with summaries",
    "outputName": "youtube_videos",
    "promptInstruction": "Find the 5 most relevant YouTube videos for this keyword or topic: [VARIABLE1]. Return them as a ranked list with video titles, URLs, and brief summaries."
  },
  {
    "taskID": 216,
    "semanticTitleOfTask": "Return final HTML Selector File",
    "taskDescription": "This skill inputs a collection of embeddable file URLs (either text, code, images, videos, audio). The skill will generate a \"selector\" HTML page which embeds these elements for quick and easy selection and review (if the files are either very large or cumbersome, it might be best to use the \"summarize\" skill instead).",
    "inputDescription": "list of files or text to return along with a description of them",
    "inputRequired": [
      "edited_html",
      "youtube_videos"
    ],
    "outputDescription": "URL of selector file",  
    "outputName": "selector_html_url",
    "promptInstruction": "Generate an HTML selector page that embeds the following files: [edited_html]. Include these descriptions for each file: [youtube_videos]"
  },
  {
    "taskID": 177,
    "semanticTitleOfTask": "Write Text (Or Copy) From Researched Text",
    "taskDescription": "This skill takes a piece of researched text and uses a LLM (typically GPT4) to return text based on this research. This is very useful for a whole range of tasks, such as writing copy about a researched product, writing articles about a research topic, etc.",
    "inputDescription": "Some text input of some kind",
    "inputRequired": [
      "edited_html"
    ],
    "outputDescription": "N/A - just output the raw text",
    "outputName": "output_text", 
    "promptInstruction": "Using the following researched text, write an article on the given topic:\n\n[edited_html]"
  },
  {
    "taskID": 187,
    "semanticTitleOfTask": "Turn Script Into Voice Over MP3",
    "taskDescription": "This skill accepts a text script for a video and converts it to an MP3 voice-over file. This skill works best with scripts of less than 4000 characters (approx max 5 minutes of video), and so if a script is expected to be much longer than this, the script should be divided up (and then concatenated back together again later, using another skill which does this).",
    "inputDescription": "Text Script",
    "inputRequired": [
      "output_text"
    ],
    "outputDescription": "MP3 File URL with both internal and external link",
    "outputName": "mp3_url",
    "promptInstruction": "Convert the following video script into a voice-over MP3 file:\n\n[output_text]"
  },
  {
    "taskID": 189,
    "semanticTitleOfTask": "Generate Text Slide Video From MP3 & Script/Transcription",
    "taskDescription": "This skill will input both an MP3 file and script to generate a text slide video (in 16:9 traditional YouTube aspect ratio).",
    "inputDescription": "MP3 URL + SCRIPT",
    "inputRequired": [
      "mp3_url",
      "output_text"
    ],
    "outputDescription": "video saved to folder with internal + external paths",
    "outputName": "video_url",
    "promptInstruction": "Generate a 16:9 text slide video using the provided MP3 file and script. MP3 URL: [mp3_url], Script: [output_text]"
  },
  {
    "taskID": 217,
    "semanticTitleOfTask": "Select Most Relevant Template To Use In Next Prompt(s)",
    "taskDescription": "This skill will input a series of templates, and select the most relevant one based on a given description of the required task.",
    "inputDescription": "description of the task",
    "inputRequired": [
      "VARIABLE1",
      "VARIABLE2"
    ],
    "outputDescription": "template (text, code.. some non-binary text template output of some kind)",
    "outputName": "selected_template",
    "promptInstruction": "Given the task description: [VARIABLE1], select the most relevant template from the following options:\n\n[VARIABLE2]"
  },
  {
    "taskID": 205,
    "semanticTitleOfTask": "Build Software Tool For Any Niche/Topic",
    "taskDescription": "This skill inputs a keyword/topic/product description and generates a GPT powered software tool which the user can sell/launch/etc.",
    "inputDescription": "keyword/topic/product description to build all variables from",
    "inputRequired": [
      "VARIABLE1"  
    ],
    "outputDescription": "URL with software link and description of what it does",
    "outputName": "software_tool_url",
    "promptInstruction": "Build a GPT-powered software tool for the following niche/topic: [VARIABLE1]. The tool should allow users to interact with it and provide valuable functionality related to the niche."
  },
  {
    "taskID": 206,
    "semanticTitleOfTask": "Generate 6x Sound Effects From Text Prompt",
    "taskDescription": "The skill generates 6 short sound effects (15 seconds long) from text prompts, using AI. For example \"sound of a monkey in the jungle\" or \"loud football stadium during sports day\". For actual melodic sounds, the \"melody\" skill should be used instead.",
    "inputDescription": "description of scene or line from video (up to 6)",
    "inputRequired": [
      "VARIABLE1"  
    ],
    "outputDescription": "URLs of 6 sound effect files with prompts",
    "outputName": "sound_effect_urls",
    "promptInstruction": "Generate 6 sound effects, each 15 seconds long, based on the following text prompts:\n\n[VARIABLE1]\n\nReturn the sound effects as a numbered list with the original prompt and links to the audio files."
  },
  {
    "taskID": 215,
    "semanticTitleOfTask": "Summarise and Return All Final Output Files",
    "taskDescription": "This skill returns a summary of various previous skills that have completed, and is particularly useful when appearing at the end of an entire Agent task, silo or sequence of skills.",
    "inputDescription": "at least one output (perhaps other supplementary ones that may be helpful)",
    "inputRequired": [
      "selector_html_url",
      "video_url",
      "software_tool_url",
      "sound_effect_urls"
    ],
    "outputDescription": "nice text summary containing 1 or more outputs",
    "outputName": "output_summary",
    "promptInstruction": "Summarize the following outputs from previous skills:\n\n[selector_html_url]\n[video_url]\n[software_tool_url]\n[sound_effect_urls]\n\nInclude key details and links to the final output files."
  },
  {
    "taskID": 216, 
    "semanticTitleOfTask": "Return final HTML Selector File",
    "taskDescription": "This skill inputs a collection of embeddable file URLs (either text, code, images, videos, audio). The skill will generate a \"selector\" HTML page which embeds these elements for quick and easy selection and review (if the files are either very large or cumbersome, it might be best to use the \"summarize\" skill instead).",
    "inputDescription": "list of files or text to return along with a description of them",
    "inputRequired": [
      "selector_html_url",
      "output_summary" 
    ],
    "outputDescription": "URL of selector file",
    "outputName": "final_selector_html_url", 
    "promptInstruction": "Generate an HTML page that embeds and allows selection of the following files:\n\n[selector_html_url]\n\nInclude these descriptions for each file:\n\n[output_summary]"
  },
  {
    "taskID": 215,
    "semanticTitleOfTask": "Summarise and Return All Final Output Files",
    "taskDescription": "This skill returns a summary of various previous skills that have completed, and is particularly useful when appearing at the end of an entire Agent task, silo or sequence of skills.",
    "inputDescription": "at least one output (perhaps other supplementary ones that may be helpful)", 
    "inputRequired": [
      "output_summary",
      "final_selector_html_url"
    ],
    "outputDescription": "nice text summary containing 1 or more outputs",
    "outputName": "final_output_summary",
    "promptInstruction": "Summarize the following outputs from the completed skills in this agent task:\n\n[output_summary]\n[final_selector_html_url]\n\nInclude key details, insights, and links to the final output files."
  },
  {
    "taskID": 230,
    "semanticTitleOfTask": "Publish Content To RSS Feed",
    "taskDescription": "Publishes a new URL to RSS feed.",
    "inputDescription": "URL to publish",
    "inputRequired": [
      "final_selector_html_url"  
    ],
    "outputDescription": "URL of RSS feed",
    "outputName": "rss_feed_url",
    "promptInstruction": "Publish the content at [final_selector_html_url] to the RSS feed and return the updated feed URL."
  }
]

Happy? Now Copy-Paste To Proceed...

BACK TO SUBAGENTS

COPY TO CLIPBOARD