What Shall We Build Next?

Describe

Describe your task

Refine

Refine the plan

SubAgents

Review all agents

Deploy

Deploy your agent

A) SUBAGENT SUMMARY: The Audio Extraction and Transcription Agent extracts audio from a video file and generates a text transcript with timestamps.

B) FINAL TASK OUTPUT: A complete, accurate text transcript of the video's audio, including timestamps for each spoken sentence or phrase, saved as a plain text file (e.g., transcription.txt).

C) SUBAGENT INPUT:
- The original "how-to" video file (e.g., "how_to_make_software_mockups_with_gimp.mp4")

E) SUBAGENT TASK SUMMARY:
1. Extract audio from the input video file using skill #191 (Extract MP3 Audio From MP4 File)
   - Input: The original video file (e.g., "how_to_make_software_mockups_with_gimp.mp4")
   - Output: The extracted audio as an MP3 file (e.g., "audio_from_video.mp3")

2. Generate a transcription of the audio with timestamps using skill #188 (Get Transcription Of MP3 (With Timings))
   - Input: The extracted audio file (e.g., "audio_from_video.mp3")
   - Output: A text file containing the transcription with timestamps (e.g., "transcription_with_timestamps.txt")

3. Save the transcription text file using skill #168 (Save text/html to server)
   - Input: The text file containing the transcription with timestamps (e.g., "transcription_with_timestamps.txt")
   - Output: The saved transcription file with its URL (e.g., "transcription.txt")

Subagent final input: The original "how-to" video file
Subagent final output: The saved transcription file with its URL

F) SILOS:
This subagent does not require any silos as the tasks can be completed sequentially without the need for parallel processing or multiple iterations.

SubAgent #1 - Diagram

Expand Diagram

A) SUBAGENT SUMMARY: This subagent analyzes the video transcription to identify the six key steps, extracts corresponding thumbnail images, and resizes and saves the images.

B) FINAL TASK OUTPUT: A list of the six main steps extracted from the video transcription, saved as a plain text file (steps.txt), and six high-resolution screenshots (1280x720 pixels) in PNG format, corresponding to the key steps in the process, saved as step1.png, step2.png, ..., step6.png.

C) SUBAGENT INPUT:
- Complete video transcription with timestamps (transcription.txt)
- URL of the original video file (video.mp4)

E) SUBAGENT TASK SUMMARY:
1. transcription.txt > #178 - Rewrite Text Based On Instructions (Analyze transcription to identify the six key steps and save the output as steps.txt) > steps.txt
2. video.mp4 + steps.txt > #202 - Extract Thumbnail Images Of MP4 Video (Extract one thumbnail image for each of the six steps identified in steps.txt) > step1_raw.png, step2_raw.png, ..., step6_raw.png
3. step1_raw.png, step2_raw.png, ..., step6_raw.png > #228 - Resize Image (Resize each thumbnail image to 1280x720 pixels) > step1.png, step2.png, ..., step6.png
4. steps.txt, step1.png, step2.png, ..., step6.png > #215 - Summarise and Return All Final Output Files (Summarize the six steps and the corresponding thumbnail images) > subagent_output_summary.txt

F) SILOS:
Silo 1: Step Identification
- transcription.txt > #178 - Rewrite Text Based On Instructions > steps.txt

Silo 2: Image Extraction and Resizing
- video.mp4 + steps.txt > #202 - Extract Thumbnail Images Of MP4 Video > step1_raw.png, step2_raw.png, ..., step6_raw.png
- step1_raw.png, step2_raw.png, ..., step6_raw.png > #228 - Resize Image > step1.png, step2.png, ..., step6.png

Silo 3: Output Summary
- steps.txt, step1.png, step2.png, ..., step6.png > #215 - Summarise and Return All Final Output Files > subagent_output_summary.txt

SubAgent #2 - Diagram

Expand Diagram

A) SUBAGENT SUMMARY: The Article Generation Agent generates the main text of an article, including an introduction, detailed explanations of six steps with corresponding image filenames, and a conclusion, based on a step-by-step breakdown and transcription.

B) FINAL TASK OUTPUT: The final output will be a plain text file (article.txt) containing the main text of the article, which includes:
- An introduction (100-200 words)
- Six main steps, each with a descriptive heading, a detailed explanation based on the video transcription (100-200 words per step), and the corresponding image filename
- A conclusion (100-200 words)

C) SUBAGENT INPUT:
- A step-by-step breakdown of the six main steps, extracted from the video transcription, in a structured format (e.g., JSON or plain text)
- The complete, accurate text transcript of the video's audio, including timestamps for each spoken sentence or phrase

E) SUBAGENT TASK SUMMARY:
1. Extract content of the step-by-step breakdown file (#236)
2. Extract content of the transcription file (#236)
3. Research the topic lightly based on the transcription and step-by-step breakdown to gather additional context (#182)
4. Generate an introduction for the article based on the research, transcription, and step-by-step breakdown (#178)
5. For each of the six steps:
   a. Generate a descriptive heading based on the step information (#178)
   b. Generate a detailed explanation of the step based on the transcription and research (#178)
   c. Extract the corresponding image filename from the step information
6. Generate a conclusion for the article based on the research, transcription, and step-by-step breakdown (#178)
7. Combine the generated introduction, six step explanations (with headings and image filenames), and conclusion into a single plain text file (#220)
8. Save the combined article text as "article.txt" on the server (#168)

F) SILOS:
This subagent does not require silos as the tasks are sequential and interdependent. However, if needed, the six steps could be processed in parallel silos to improve efficiency:

Silo 1: Generate introduction
Silo 2: Generate step 1 explanation
Silo 3: Generate step 2 explanation
Silo 4: Generate step 3 explanation
Silo 5: Generate step 4 explanation
Silo 6: Generate step 5 explanation
Silo 7: Generate step 6 explanation
Silo 8: Generate conclusion

After all silos are complete, combine the generated text and save the final article.

SubAgent #3 - Diagram

Expand Flow

A) SUBAGENT SUMMARY: The Formatting and Assembly Agent takes the generated article text and extracted images, then formats them into a visually appealing and properly structured HTML file, which is saved and ready for publication.

B) FINAL TASK OUTPUT: A well-formatted, publication-ready HTML file containing the generated article text with the extracted images embedded at appropriate locations within the content. The HTML file will be saved on the server, with both internal and external URLs provided for access.

C) SUBAGENT INPUT:
1. Plain text file containing the generated article text (introduction, six main steps with headings, and conclusion)
2. Six high-resolution image files (PNG or JPG) corresponding to the key steps in the article
3. A JSON file or similar structured data specifying the image filenames and their corresponding step numbers

E) SUBAGENT TASK SUMMARY:
1. Load and extract content from the plain text article file (#236)
2. Rewrite the extracted article text into properly structured HTML, with appropriate heading tags (h1, h2, etc.), paragraph tags, and other necessary HTML elements (#179)
3. Load and extract image filenames and step numbers from the JSON file (#236)
4. For each image:
   a. Load the image file (#174)
   b. Resize the image to a web-friendly size, if necessary (#228)
   c. Generate an <img> tag with the appropriate 'src' attribute and 'alt' text (#179)
   d. Embed the <img> tag into the corresponding step section within the HTML article (#179)
5. Save the final formatted HTML file to the server (#168)
6. Return the internal and external URLs of the saved HTML file (#215)

Subagent final inputs: plain text article file, six image files, JSON file with image details
Sequence: #236 (article text) > #179 > #236 (JSON) > #174 > #228 (if needed) > #179 > #179 > #168 > #215
Subagent final outputs: internal and external URLs of the formatted HTML article

F) SILOS: This subagent does not require silos, as the process is fairly linear and does not involve multiple distinct parts that need to be processed independently. The tasks can be completed sequentially without the need for parallel processing in silos.