What Shall We Build Next?

Describe

Describe your task

Refine

Refine the plan

SubAgents

Review all agents

Deploy

Deploy your agent

A) SUBAGENT SUMMARY  
“This subagent thoroughly researches ‘smart home technology for electronic technicians’ and then generates a structured 15-module course outline (with module titles, short descriptions, and learning goals), accompanied by a concise research summary.”

────────────────────────────────────────────────────────
B) FINAL TASK OUTPUT  
A single textual output (e.g., a JSON or text file) containing:  
1) A 1000–3000 character research summary on “smart home technology for electronic technicians.”  
2) A detailed 15-module course outline with module titles, short descriptions, and learning goals.  
(Collectively called [research-and-module-outline].)

────────────────────────────────────────────────────────
C) SUBAGENT INPUT  
• Primary topic or keywords to research (e.g., “Smart home technology for electronic technicians”).

────────────────────────────────────────────────────────
E) SUBAGENT TASK SUMMARY

1) Subagent receives the initial input:  
   • “Smart home technology for electronic technicians” (or similar keywords).

2) Execute Skill #216 - Research Topic (Or Keyword) Deeply  
   • Input: “Smart home technology for electronic technicians.”  
   • Action: Conducts extensive online research, summarizes key standards, trends, training requirements, best practices, etc.  
   • Output: A 1000–3000 character text summary of the researched info.

3) Execute Skill #223 - Powerful LLM Prompt-to-Text Response  
   • Input(s):  
       a) The research summary from Skill #216.  
       b) A prompt such as:  
          “Create a 15-module outline for a smart home course aimed at electronic technicians. Include module titles, short descriptions, and learning goals in a structured format.”  
   • Action: Generates a structured outline (text or JSON) containing 15 modules, each with a title, brief description, and learning objectives.  
   • Output: 15-module structured course outline.

4) Subagent Final Output ([research-and-module-outline])  
   • Combine the research summary output from Step 2 and the 15-module outline from Step 3 into a single result.  
   • Format the output so that it can be easily passed to the next subagent in the overall workflow.

────────────────────────────────────────────────────────
F) SILOS

• SILO 1: RESEARCH  
  - Goal: Collect deep research on the assigned topic.  
  - Skill Used: #216

• SILO 2: CREATE MODULE OUTLINE  
  - Goal: Convert the research into a 15-module outline (titles, descriptions, goals).  
  - Skill Used: #223

Each silo feeds into the next, culminating in the final combined text output.

SubAgent #1 - Diagram

Expand Diagram

A) SUBAGENT SUMMARY  
This subagent’s purpose is to generate the instructional text content for each of the 15 modules in the Smart Home course, tailoring each module’s text to electronic technicians in a thorough, 2–5 page format.

B) FINAL TASK OUTPUT  
• 15 separate text files (or text blocks), each containing 2–5 pages of instructional content for one module.

C) SUBAGENT INPUT  
• A module outline containing the 15 module topics, along with any short descriptions and learning objectives (output from the Module Outline & Research Agent).

D) SUBAGENT TASK SUMMARY  
1) Receive the module outline (includes module titles, short descriptions, objectives).  
2) For each module:  
   a) Use Skill #223 (Powerful LLM Prompt-to-Text Response) to generate ~2–5 pages of instructional content specifically aimed at electronic technicians.  
   b) Store the resulting  text for that module.  
3) Compile all 15 text outputs into a final collection of 15 module texts.

────────────────────────────────  
   Flow Overview:  
   Subagent input (module outline)  
          ▼  
   Repeat for each of the 15 modules → #223 - Powerful LLM Prompt-to-Text Response  
          ▼  
   Subagent final output: [module-text-content] (15 text files/blocks)

E) SILOS  
• Silo 1 – “Generate Content for Module 1”  
  – Input: Module outline entry for Module 1  
  – Action: #223 (Prompt-to-Text) to produce 2–5 pages of text  
  – Output: text file for Module 1

• Silo 2 – “Generate Content for Module 2”  
  – Repeat the same approach using #223, output text for Module 2

• … (Continue similarly for Modules 3 through 15)

• Final Silo Output  
  – Gather all 15 module text files (one per module) as an array/list of text files. This becomes the subagent’s final output.

SubAgent #2 - Diagram

Expand Diagram

A) SUBAGENT SUMMARY  
This subagent takes the final text content for each Smart Home course module and generates a 15-question quiz (with correct answers) for each module.

B) FINAL TASK OUTPUT  
A set of quiz texts (15 total—one per module) containing 15 questions (mixed multiple choice, true/false, short answer) and an answer key for each module.

C) SUBAGENT INPUT  
• The final text content of each module (e.g., the detailed instructional text).  
• The module’s topic/title (e.g., “Networking Protocols in Smart Homes,” etc.).

D) SUBAGENT TASK SUMMARY  
1) Receive the text content for each module (and identify the module's topic).  
2) For each module, call Skill #223 (“Powerful LLM Prompt-to-Text Response”) with a prompt that instructs it to generate a 15-question quiz, including the correct answers.  
   – The prompt can be structured as:  
     "From the following module text, create a 15-question quiz for [module topic]. Use multiple-choice, true/false, and short-answer questions. Provide a correct answer key. The content should be suitable for electronic technicians learning smart home technology."  
3) The output from Skill #223 for each module is a text document containing the 15 questions and an answer key.  
4) Return or store these 15 quiz texts as the subagent’s final output.

E) SILOS  
Since the only core function here is to generate quizzes, we can think of each module quiz creation as a single repeated silo:

• Silo: “Generate quiz for Module X”  
  1) Input: Module X text + Module X topic  
  2) Action:  
     – Skill #223 → “Create a 15-question quiz (multiple choice, true/false, short answer) with correct answers for [module topic].”  
  3) Output: Single text block => [Quiz questions + answer key for Module X].

Repeat the above silo for each of the 15 modules, yielding 15 separate quiz text outputs in total.

SubAgent #3 - Diagram

Expand Flow

A) SUBAGENT SUMMARY  
This subagent’s core purpose is to generate (or optionally find) relevant images for each module of the Smart Home course, so that these images can later be inserted into the final PDF documents.

B) FINAL TASK OUTPUT  
A structured set of image URLs (most often PNG files) for each module. Typically returned as an array or list of URLs for each module, for example:  
[
  {
    "moduleNumber": 1,
    "imageURLs": ["https://server.com/…png", "…"]
  },
  {
    "moduleNumber": 2,
    "imageURLs": ["https://server.com/…png", …]
  },
  …
]

C) SUBAGENT INPUT  
• Module names or titles  
• Brief description of each module’s content/focus  
• (Optionally) number of images desired per module and any stylistic considerations

E) SUBAGENT TASK SUMMARY  
Below is a typical step-by-step flow of how the Image Generation Agent can operate for each module. In practice, these steps may be repeated once per module (i.e., repeated 15 times overall if all modules need images).

1) RECEIVE MODULE INFO  
   - Input: "Module name/title," “Brief description,” and optionally “Number of images needed,” “Style guidelines,” etc.

2) (OPTIONAL) GENERATE IMAGE PROMPT TEXT (Skill #223 – Powerful LLM Prompt-to-Text Response)  
   - Purpose: If a more elaborate or stylistically consistent image prompt is needed, the subagent can call #223 to transform the module description into a creative, well-structured image prompt.  
   - Input Example:  
     {
       "prompt": "Generate a concise prompt for an AI image that illustrates the concepts of a smart security system aimed at electronic technicians. The image should have a modern feel and highlight sensors and cameras."
     }  
   - Output: A short text prompt suitable for the image generator (e.g. “Digital illustration of a modern smart home security layout with labeled sensors, cameras, and a central control panel”).

3) CREATE AI IMAGE (Skill #222 – Make Image (Especially With Text))  
   - Purpose: Generate the actual image(s) from the prompt.  
   - Input: The final text prompt (either user-provided or from Step 2).  
   - Output: One or more PNG image URLs. Each image (1024x1024 or as supported) is automatically saved and returned.

4) (OPTIONAL) FIND ADDITIONAL/ALTERNATIVE IMAGES (Skill #187 – Find Relevant Images with Brave Search)  
   - Purpose: If the user also wants to include real-life photos, diagrams, or references from the Internet, the subagent can optionally do a web image search.  
   - Input Example:  
     {
       "keyword": "smart home security system blueprint schematic"
     }  
   - Output: Up to 5 relevant public image URLs from across the web.

5) RETURN IMAGE URLS  
   - Combine results from Step 3 (and optionally Step 4), producing the final subagent output for each module.  
   - Final Subagent Output: [module-images], typically in JSON array form, containing all the new AI-generated URLs and/or any curated images from search.

F) SILOS  
• SILO 1: (Optional) Prompt Generation  
   - Step 2 (#223) – Convert module’s text into an ideal AI image prompt if needed.

• SILO 2: AI-Based Image Creation  
   - Step 3 (#222) – Create one or more AI images using the prompt from SILO 1 or from direct user input.

• SILO 3: Web-Based Image Search (Optional)  
   - Step 4 (#187) – Retrieve up to 5 relevant publicly available images from the Internet.

• SILO 4: Collation & Output  
   - Step 5 – Consolidate AI-generated and/or found images into a stable output format (list of image URLs), then return them.

4 Template & Links

Expand Flow

A) SUBAGENT SUMMARY:  
This subagent assembles each module’s text, quiz, and images into a final PDF. Because there is no built-in PDF creation skill here, it will compile the finished course content (text + quiz + images) into an intermediate format (e.g., HTML or Markdown) and then pass that to an external PDF creation tool/script.

B) FINAL TASK OUTPUT:  
15 separate PDF files (one per module), each containing:  
1) The module’s instructional text.  
2) Embedded images (or image references).  
3) A 15-question quiz appended at the end.

C) SUBAGENT INPUT:  
• An array/list of 15 module content bundles, where each bundle includes:  
  – “moduleText” (the module’s instructional text).  
  – “moduleQuiz” (the module’s 15 questions + answer key).  
  – “moduleImages” (any relevant image URLs for that module).

E) SUBAGENT TASK SUMMARY:

The assembly subagent performs these steps for each of the 15 modules:

1) Collect Data For Module  
   – Accept “moduleText” (string), “moduleQuiz” (string), and “moduleImages” (array of URLs) as input.

2) Compile Text & Images In An Intermediate Format  
   – (Optional) Use Skill #223 (“Powerful LLM Prompt-to-Text Response”) to transform or finalize the layout in HTML or Markdown, if needed. For instance:  
     Input Prompt Example:  
       “Here is the text for this module and the quiz. Embed these image URLs as <img> tags or references in a short HTML layout. Return the compiled HTML so it can be converted into a PDF.”  
     Output Example:  
       “<html>  
         <body>  
           <h1>Module Title</h1>  
           <p>[Module text here]</p>  
           <img src='[Image1]' />  
           <img src='[Image2]' />  
           …  
           <h2>Quiz</h2>  
           [Quiz questions + answers appended]  
         </body>  
       </html>”

3) Generate The PDF (External Tool)  
   – Pass the compiled HTML/Markdown document + images to the external “PDF creation/conversion tool” (not currently in the skill list).  
   – Receive back a final PDF file for this module.

4) Output The Final PDF For Each Module  
   – Collect the URLs or file references of all 15 PDFs, each one representing a finished module.

F) SILOS:

• Silo #1: Data Compilation & Formatting  
  - For each module, gather text, quiz, and image URLs.  
  - (Optional) Format them into a structured layout (HTML, Markdown, or similar) using skill #223 if any transformation is required.

• Silo #2: PDF Generation  
  - In an external step (outside the listed skill set), call a PDF generation script/tool to convert the compiled content into a PDF.  
  - Return the final PDF file reference.

END OF SUBAGENT FLOW