What Shall We Build Next?

Describe

Describe your task

Refine

Refine the plan

SubAgents

Review all agents

Deploy

Deploy your agent

A) SUBAGENT SUMMARY:  
"WP PLUGIN BUILDER" is responsible for generating the plugin.php file that registers and enqueues the chatbot’s assets and integrates the chatbot into WordPress.

B) FINAL TASK OUTPUT:  
A fully formed “plugin.php” file (in plain text) that includes:  
• A valid WordPress plugin header (with name, version, author, etc.)  
• Enqueueing of chatbot.js, chatbot.css, and reference to config.json  
• Mechanism (footer hook or shortcode) to display the chatbot container

C) SUBAGENT INPUT:  
• Plugin metadata (name, version, author, description)  
• File references for chatbot.js, chatbot.css, and config.json (i.e. file paths or instructions on how to load them)  
• Desired integration method (action hook for footer or shortcode for insertion on pages)

E) SUBAGENT TASK SUMMARY:  
1) Take the subagent input (plugin metadata + instructions).  
2) Call Skill #223 (Powerful LLM Prompt-to-Text Response) with a detailed prompt that includes:  
   – Mandated WordPress plugin header format and metadata.  
   – Instructions to enqueue chatbot.js, chatbot.css, and reference config.json.  
   – Instructions to generate either a footer hook or a shortcode function for displaying the chatbot container.  
3) Receive the generated plugin.php code from Skill #223.  
4) Output final plugin.php text.

F) SILOS:

» SILO 1: PREPARE PROMPT  
   – Gather the plugin name, description, version, author, and any special instructions about how to register/enqueue scripts in WordPress.  
   – Ensure all references (chatbot.js, chatbot.css, config.json) are included.

» SILO 2: GENERATE PLUGIN.PHP (Using #223)  
   – Input: Detailed prompt from SILO 1 (plugin meta, requirements, code structure).  
   – Skill: #223 – Powerful LLM Prompt-to-Text Response.  
   – Output: Completed plugin.php text that:  
       1) Registers plugin details.  
       2) Properly enqueues JS/CSS when WordPress loads.  
       3) Optionally references config.json for data.  
       4) Implements a function or shortcode to insert the chatbot container in the page footer or where desired.

» SILO 3: RETURN FINAL FILE  
   – The completed plugin.php code is returned as the subagent’s final output.

SubAgent #1 - Diagram

Expand Diagram

A) SUBAGENT SUMMARY:  
This subagent will generate a single JavaScript file ("chatbot.js") containing all the logic for a minimal chat widget, including opening/closing the chat UI, capturing user queries, calling an AI endpoint or referencing local site data (e.g., config.json), and displaying responses.

B) FINAL TASK OUTPUT:  
• A fully-formed JavaScript file (chatbot.js) as plain text.

C) SUBAGENT INPUT:  
• A detailed text prompt or specification describing the required chatbot functionality, UI flow, event handling, placeholders for AI endpoint calls, references to config.json, and any other desired customizations.

E) SUBAGENT TASK SUMMARY:  
1) Start: Subagent receives the specification of the chatbot features (initial input).  
2) Skill #223 – Powerful LLM Prompt-to-Text Response:  
   a) Provide #223 with an instruction prompt (constructed from the subagent’s input) to generate a minimal but functional chatbot.js.  
   b) Specify that the code should do the following:  
      • Create/append relevant HTML elements for the chat window (e.g., chat container, message list, input box, send button).  
      • Implement open/close logic for a chat widget toggle icon.  
      • Capture user input and either call a placeholder AI endpoint or pull relevant responses from config.json.  
      • Display AI responses in the chat window.  
      • Include basic error handling or fallback.  
   c) Receive the complete JavaScript code as text output.  
3) Final Output: Return the generated chatbot.js code as the subagent’s result.

F) SILOS:  
• Silo 1: Prompt Preparation  
  - The subagent prepares a template prompt to define the chatbot’s requirements clearly.

• Silo 2: Code Generation  
  - Pass the prompt to skill #223 and obtain the complete JS file text.

• Silo 3: Final Output  
  - Return the final chatbot.js (plain text).

SubAgent #2 - Diagram

Expand Diagram

A) SUBAGENT SUMMARY  
This subagent ("CHATBOT STYLIST") is responsible for creating the chatbot.css file, which defines all layout, positioning, colors, fonts, transitions, and responsiveness for the chatbot widget.

B) FINAL TASK OUTPUT  
The final output of this subagent is a single text file named “chatbot.css” containing valid, production-ready CSS code for the chatbot widget.

C) SUBAGENT INPUT  
• A textual description of desired chatbot styling, including:  
  – Chat window size, color scheme, positioning (desktop/mobile), transitions/animations.  
  – Button/icon styling, hover states, or custom fonts.  
  – Overall responsiveness requirements (how the widget behaves on smaller screens).

E) SUBAGENT TASK SUMMARY

1) Receive Chatbot Styling Specifications  
   – The subagent starts by collecting the detailed style requirements (position, color, transitions, fonts, responsiveness) from the main workflow or from user-provided instructions.

2) Generate Proposed CSS Code (Skill #223 – Powerful LLM Prompt-to-Text Response)  
   – Input: A prompt that includes the gathered style requirements and the request to produce minimal, responsive CSS.  
   – Action: The LLM (#223) is instructed to create efficient, well-structured CSS for the chatbot.  
   – Output: The raw CSS code in text form.

3) Return Final chatbot.css  
   – The subagent returns the text output as the final artifact.

F) SILOS  
There is only one primary “silo” for this subagent—the generation of the CSS code itself. The steps are straightforward:  
• Silo 1: Gather styling specs → Use skill #223 to create CSS → Output chatbot.css.

That completes the workflow for the “CHATBOT STYLIST” subagent.

SubAgent #3 - Diagram

Expand Flow

A) SUBAGENT SUMMARY  
This subagent’s purpose is to generate a German Shepherd–themed chatbot icon with a transparent background and optionally resize it to the desired final dimensions.

B) FINAL TASK OUTPUT  
A single PNG file (e.g., shepherd-icon.png) with a transparent background featuring a friendly German Shepherd dog head, sized appropriately for use in the chatbot widget.

C) SUBAGENT INPUT  
• Icon prompt text (e.g., “friendly German Shepherd dog head, simple silhouette, transparent background”)  
• Final icon dimensions, if resizing is required (e.g., 64×64)

E) SUBAGENT TASK SUMMARY  
1) Receive the subagent inputs:  
   - A text prompt describing the desired icon style and subject.  
   - (Optional) a target dimension for the final icon (width × height in pixels).

2) USE SKILL #182 – Create Dalle Image (Transparent Square):  
   - INPUT: The icon prompt text.  
   - OUTPUT: A 1024×1024 square PNG image URL (transparent background).

3) (Optional) If a smaller icon is needed, USE SKILL #191 – Resize Image:  
   - INPUT: The 1024×1024 PNG URL from step 2, plus the desired final width × height (e.g., 64×64).  
   - OUTPUT: Resized PNG image URL (transparent background).

4) Provide the final PNG image URL (which becomes shepherd-icon.png) as the subagent output.

F) SILOS  
• Silo 1: ICON CREATION  
   - Single skill call (#182) to generate a 1024×1024 PNG.

• Silo 2: ICON RESIZING (Optional)  
   - Single skill call (#191) to resize the PNG to the specified final dimensions.

• Final Output: shepherd-icon.png (or an equivalent PNG file URL)

4 Template & Links

Expand Flow

A) SUBAGENT SUMMARY  
──────────────────────────────────────────────────────  
“CONFIG DATA CURATOR” is responsible for gathering relevant facts about ShepherdDigest.com’s site content and multiple shepherd dog breeds, then producing a valid config.json file that references these findings and includes placeholders for an AI API key.

B) FINAL TASK OUTPUT  
──────────────────────────────────────────────────────  
The final output of this subagent is a single text file named "config.json" containing:  
• Summaries of site content (ShepherdDigest.com)  
• Trivia/facts about various shepherd dog breeds (German Shepherd, Belgian Malinois, Australian Shepherd, etc.)  
• AI configuration placeholders (e.g., "apiKey": "YOUR_AI_KEY_HERE").

C) SUBAGENT INPUT  
──────────────────────────────────────────────────────  
1) The domain or reference to ShepherdDigest.com (so it knows what site content to reference).  
2) A list of shepherd dog breeds to include (e.g., “German Shepherd,” “Belgian Malinois,” “Australian Shepherd,” etc.).  
3) Any instructions about how the JSON should be structured, along with placeholders for the plugin’s AI key.

E) SUBAGENT TASK SUMMARY (Step-by-Step Use of Skills)  
──────────────────────────────────────────────────────

1) GATHER SITE CONTENT & BREED DATA  
   • Use Skill #216 – “Research Topic (Or Keyword) Deeply”  
     INPUT: A combined prompt referencing:  
       • ShepherdDigest.com content (so it can gather an overview of the site’s key pages/features)  
       • The shepherd dog breeds in scope (e.g., German Shepherd, Belgian Malinois, Australian Shepherd, etc.)  
     OUTPUT: A text “research summary” containing:  
       – Key points from ShepherdDigest.com (major topics, articles, pages, or any relevant site highlights)  
       – Basic facts and trivia about each identified shepherd breed.

2) PREPARE THE CONFIG.JSON  
   • Use Skill #223 – “Powerful LLM Prompt-to-Text Response”  
     INPUT:  
       – The “research summary” from step 1.  
       – A concise instruction: “Format the above information into valid JSON. Include sections for site references, breed facts, and a placeholder ‘apiKey’ field for the plugin. Make sure the JSON keys are meaningful (e.g., ‘siteContent’, ‘breedFacts’, ‘apiKey’). Ensure the JSON is syntactically valid.”  
     OUTPUT:  
       – The complete JSON text suitable to be saved as config.json.

F) SILOS (Logical Grouping of Subtasks)  
──────────────────────────────────────────────────────

• SILO 1: RESEARCH  
  – Step 1: #216 (Research ShepherdDigest.com + shepherd breeds) → Output: “Research Summary Text”

• SILO 2: JSON CREATION  
  – Step 2: #223 (Convert research summary to final config.json structure with placeholders) → Output: “config.json” text

──────────────────────────────────────────────────────

This completes the “CONFIG DATA CURATOR” workflow.

5 Template & Links

Expand Flow

A) SUBAGENT SUMMARY  
"PLUGIN PACKAGER" will take all generated files (plugin.php, chatbot.js, chatbot.css, shepherd-icon.png, config.json) produced by the other subagents, then bundle them into a single .zip file for distribution.

B) FINAL TASK OUTPUT  
The final output is a single “chatbot-plugin.zip” file containing:  
1. plugin.php  
2. chatbot.js  
3. chatbot.css  
4. shepherd-icon.png  
5. config.json

C) SUBAGENT INPUT  
A list/array of file references pointing to each of the five required files (i.e., their server paths/URLs or direct references to the in-memory file content).

E) SUBAGENT TASK SUMMARY  
1) Receive file references (plugin.php, chatbot.js, chatbot.css, shepherd-icon.png, config.json) from earlier subagents.  
2) Invoke a new skill or process (not in current skill list) to bundle multiple files into a single zip archive (e.g., skill #300: “CreateZipArchiveOfMultipleFiles”).  
   • INPUT to #300: array of (fileName, fileContent/fileURL).  
   • OUTPUT from #300: chatbot-plugin.zip  
3) Return the final chatbot-plugin.zip as the subagent’s output.

F) SILOS  
Since packaging the plugin is a relatively straightforward step, there is effectively only one silo here:  
• SILO 1: Packaging  
   1) Collect the 5 files.  
   2) Zip them using the (to-be-added) packaging skill.  
   3) Return chatbot-plugin.zip.