Promo Video from Script Input - 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": 190,
    "semanticTitleOfTask": "Generate Timed Promotional Script Content",
    "taskDescription": "This task uses skill #190 to produce a short promotional script with approximate timings for a hook (0–4s), a CTA (5–9s), and hashtags (10–14s), ensuring a total duration of 10–15 seconds.",
    "inputDescription": "Accepts two inputs: (1) VARIABLE1, which indicates the subject or topic, and (2) VARIABLE2, which are any specific user directives (e.g., style, tone, or special requirements).",
    "inputRequired": [
      "VARIABLE1",
      "VARIABLE2"
    ],
    "outputDescription": "A concise promotional script containing timed segments for the hook, CTA, and hashtags, formatted to show each segment with clear start and end times.",
    "outputName": "short_promotional_script",
    "promptInstruction": "You are a professional social media script writer. Create a short promotional script for 'VARIABLE1' within 10–15 seconds. Requirements: • HOOK (0–4s): up to 12 words, capturing attention. • CTA (5–9s): up to 10 words, driving an action. • HASHTAGS (10–14s): 3–4 relevant tags. Use casual, conversational language, clearly labeling each timed segment. Incorporate any 'VARIABLE2' provided."
  },
  {
    "taskID": 223,
    "semanticTitleOfTask": "Parse Script Keywords Using LLM for B-roll search",
    "taskDescription": "This skill uses a large language model to analyze the short promotional script and extract key search terms. The output is a short list (3–5) of relevant keywords that reflect the script’s theme or subject matter, to facilitate a targeted B-roll search.",
    "inputDescription": "The task requires the short_promotional_script, providing textual content that mentions the target theme, audience, or style. The LLM will parse this text to determine the most pertinent keywords.",
    "inputRequired": [
      "short_promotional_script"
    ],
    "outputDescription": "A concise comma-separated list of 3–5 keywords that best represent the script's primary ideas or themes, used to find relevant vertical B-roll videos.",
    "outputName": "refined_keywords",
    "promptInstruction": "You are a video content keyword specialist. Analyze the provided promotional script and extract 3–5 most relevant keywords that would be effective for searching vertical video content. Focus on visual elements and themes mentioned in the script. Consider both explicit and implicit visual themes. Format your response as a simple comma-separated list of keywords. Example output format: 'nature, sunset, ocean, meditation'."
  },
  {
    "taskID": 205,
    "semanticTitleOfTask": "Retrieve Vertical Royalty-Free Video Content",
    "taskDescription": "This task uses an external API or database to fetch up to five vertical video links based on the provided keywords. The videos are intended for use as background B-roll clips in the final promotional video.",
    "inputDescription": "The input is the refined list of keywords from the preceding LLM-driven keyword extraction step. These keywords are used to search for relevant vertical-format, royalty-free video content.",
    "inputRequired": [
      "refined_keywords"
    ],
    "outputDescription": "Up to five royalty-free vertical video URLs that match the provided keywords. Each URL points to downloadable vertical video resources.",
    "outputName": "vertical_video_urls",
    "promptInstruction": "no instruction"
  },
  {
    "taskID": 223,
    "semanticTitleOfTask": "Select Best-Matching Video URL from the list",
    "taskDescription": "This task uses a language model to evaluate the retrieved video URLs and the promotional script, choosing a single best URL that aligns with the script’s themes and branding. It ensures the chosen footage is the strongest match for the intended promotional style.",
    "inputDescription": "This requires the complete list of vertical video URLs from the previous step, plus the original promotional script. The LLM compares each video’s visual characteristics to the script’s content and brand requirements, then picks a single URL.",
    "inputRequired": [
      "vertical_video_urls",
      "short_promotional_script"
    ],
    "outputDescription": "A single vertical video URL, chosen as the optimal match for the script’s narrative and visual style. Optionally includes a one-sentence justification.",
    "outputName": "selected_video_url",
    "promptInstruction": "You are a video content curator. Review these vertical video URLs and the original promotional script. Select the single best-matching video that aligns with our script's message and brand. Consider visual appeal, relevance to the script's theme, and professional quality. Format your response as a single URL with a brief one-sentence explanation of why it's the best choice."
  },
  {
    "taskID": 194,
    "semanticTitleOfTask": "Trim Chosen Video to Required Duration",
    "taskDescription": "This task uses a technical function to extract a 10–15 second snippet from the selected MP4 video, based on a chosen start time and duration. The resulting output is a vertically oriented clip suitable for the short promotional video.",
    "inputDescription": "The task requires the best-matching MP4 URL from the previous selection step. The function will download and cut the video to the desired 10–15 second segment, according to the designated start/end.",
    "inputRequired": [
      "selected_video_url"
    ],
    "outputDescription": "A single trimmed MP4 file (vertical orientation, 10–15 seconds long), ready for incorporation into the final promotional short video.",
    "outputName": "trimmed_vertical_broll",
    "promptInstruction": "no instruction"
  },
  {
    "taskID": 220,
    "semanticTitleOfTask": "Generate 3 short WAV files from style prompt",
    "taskDescription": "Takes the music-style prompt and generates 3 short WAV files (7–15 seconds each) matching the requested style.",
    "inputDescription": "The main input is a single string describing the desired music style or mood (VARIABLE3).",
    "inputRequired": [
      "VARIABLE3"
    ],
    "outputDescription": "Three short WAV audio files (each 7–15 seconds) that match the requested style prompt.",
    "outputName": "wav_file_urls",
    "promptInstruction": "no instruction"
  },
  {
    "taskID": 223,
    "semanticTitleOfTask": "Compile WAV links into final reference",
    "taskDescription": "Takes the 3 generated WAV files and creates a structured text listing of them, including URL, duration, and style match.",
    "inputDescription": "Receives the three WAV file URLs from the previous task (wav_file_urls).",
    "inputRequired": [
      "wav_file_urls"
    ],
    "outputDescription": "A concise text listing referencing each WAV file, including its URL, duration, and how it matches the requested music style.",
    "outputName": "short_bgm_wav_files",
    "promptInstruction": "You are a music asset compiler. You receive three WAV file URLs generated for the style 'VARIABLE3'. Create a structured list with: 1) File URL, 2) Duration (in seconds), and 3) A short description explaining how it matches the requested style. Keep descriptions concise, focusing on how each file fulfills the style criteria. Format your response as: Audio File 1: - URL: ... - Duration: ... - Style Match: ... Audio File 2: ... Audio File 3: ..."
  },
  {
    "taskID": 190,
    "semanticTitleOfTask": "GPT4 - Convert short script & guidelines into timed overlay instructions",
    "taskDescription": "This single-step subagent uses a text rewriting skill to transform the short_promotional_script and optional style guidelines into a structured, time-coded text overlay plan for a 10–15s video. Timings are allocated for each segment’s text (hook, CTA, hashtags).",
    "inputDescription": "Requires the short_promotional_script plus any optional text_overlay_guidelines. The script provides the raw content (hook, CTA, hashtags), while guidelines may specify style or alignment preferences.",
    "inputRequired": [
      "short_promotional_script",
      "text_overlay_guidelines"
    ],
    "outputDescription": "Generates a line-by-line overlay with approximate timestamps for each text segment, ensuring clarity, readability, and adherence to any style instructions. The final plan covers hook, call-to-action, and hashtags within a 10–15 second timeframe.",
    "outputName": "text_overlay_plan",
    "promptInstruction": "You are a video overlay specialist. Using the short_promotional_script & any guidelines, create a 10–15s time-coded text plan. Each line: [0:00-0:03]: “Text” | [Style]. Segments: Hook, CTA, Hashtags. Maintain 3–4s each. Ensure readability, alignment, and smooth transitions."
  },
  {
    "taskID": 501,
    "semanticTitleOfTask": "Select one WAV from the three generated short background music tracks",
    "taskDescription": "This task chooses one WAV file out of the three provided from subagent 3 for use in the final text-slide video generation step.",
    "inputDescription": "The input is an array of three WAV file URLs created by the Music Curator subagent, from which one file is selected.",
    "inputRequired": [
      "short_bgm_wav_files"
    ],
    "outputDescription": "A single chosen WAV file, ready to be used as background audio for the text-slide video.",
    "outputName": "chosen_wav_audio",
    "promptInstruction": "no instruction"
  },
  {
    "taskID": 201,
    "semanticTitleOfTask": "Generate text-slide MP4 with background audio & timed overlays",
    "taskDescription": "Combines the chosen WAV file and the text overlay plan to produce a short, vertical MP4 with synchronized text slides and background music.",
    "inputDescription": "Requires one chosen WAV file from the previous task and the timed text overlay instructions. These inputs are merged into a vertical 9:16 MP4.",
    "inputRequired": [
      "chosen_wav_audio",
      "text_overlay_plan"
    ],
    "outputDescription": "A 10–15 second vertical MP4 file with timed text slides and the chosen background WAV audio track.",
    "outputName": "text_slide_mp4",
    "promptInstruction": "no instruction"
  },
  {
    "taskID": 199,
    "semanticTitleOfTask": "Add Videos On Top Of Existing MP4",
    "taskDescription": "Combines two MP4 files by layering one file on top of or behind the other, preserving vertical format and synchronizing their timing into a single cohesive video output.",
    "inputDescription": "Takes two MP4 files: text_slide_mp4 serving as the base video, and trimmed_vertical_broll, which is to be placed behind or on top. Also needs basic layering/positioning instructions.",
    "inputRequired": [
      "text_slide_mp4",
      "trimmed_vertical_broll"
    ],
    "outputDescription": "Produces a merged vertical MP4, combining text slides, background music, and b-roll footage into a 10–15 second promotional short video.",
    "outputName": "final_promotional_short_mp4",
    "promptInstruction": "no instruction"
  }
]

Happy? Now Copy-Paste To Proceed...

BACK TO SUBAGENTS

COPY TO CLIPBOARD