What Shall We Build Next?

Describe

Describe your task

Refine

Refine the plan

SubAgents

Review all agents

Deploy

Deploy your agent

A) SUBAGENT SUMMARY:  
"Niche Researcher" obtains a list of 20 profitable ClickBank niches and provides short notes on each niche's competition and profitability ranking.

B) FINAL TASK OUTPUT:  
[ranked-niche-list] → A concise, ranked list of 20 ClickBank niches (1–20), each with a sentence or two about competition and profitability.

C) SUBAGENT INPUT:  
• {task-requirements} – Any instructions or details concerning how the subagent should research niches or what parameters to prioritize (e.g., gravity, earning potential, popularity trends).

E) SUBAGENT TASK SUMMARY:  
1) The subagent uses Skill #224 (Oracle Ask A Question) to query up-to-date data on the top 20 profitable ClickBank niches, focusing on gravity, popularity, and earning potential.  
2) The raw response from Skill #224 is passed into Skill #223 (Powerful LLM Prompt-to-Text Response), which formats the answer into a concise list of 20 ranked niches, each with a short note on competition and profitability (the [ranked-niche-list]).

F) SILOS:  
• Silo 1:  
   • INPUT: {task-requirements}  
   • ACTION (Skill #224 – Oracle Ask A Question): “List 20 profitable ClickBank niches by gravity, popularity, and earning potential. Provide short notes on competition and profitability for each.”  
   • ACTION (Skill #223 – Powerful LLM Prompt-to-Text Response): Format the oracle’s response into a final ordered list (1–20) and ensure each niche entry includes concise commentary on competition/profitability.  
   • OUTPUT: [ranked-niche-list]

SubAgent #1 - Diagram

Expand Diagram

A) SUBAGENT SUMMARY  
"Keyword Analyzer" takes in the list of 20 niches (from Subagent 1) plus any relevant requirements, and returns for each niche: 3–5 primary keywords, 3–5 long-tail keywords, and approximate search volume/competition data.

B) FINAL TASK OUTPUT  
[keyword-research], which is a consolidated, structured set of data for all 20 niches. Each niche entry contains:  
• 3–5 primary keywords + metrics (e.g., search volume, competition)  
• 3–5 long-tail keywords + metrics

C) SUBAGENT INPUT  
{task-requirements}, [ranked-niche-list]

D) SUBAGENT OUTPUT  
[keyword-research]

E) SUBAGENT TASK SUMMARY  
Below is the sequence of actions for Subagent 2, showing how the required inputs flow into the final output.

1) COMBINE INPUTS → PREPARE SINGLE PROMPT  
   - The agent takes [ranked-niche-list] (i.e., 20 niches with short notes on profitability/competition) and the {task-requirements} (any special instructions about metrics or formatting).  
   - The agent then composes a single request for keywords/metrics.

2) GENERATE KEYWORDS & BASIC METRICS (#218 OR #224)  
   • Primary Skill: #218 – Brainstorm Related Keywords  
     - DESCRIPTION: Accepts a text prompt and returns up to 10 relevant keywords with commentary.  
     - The subagent either loops over the 20 niches individually or makes one combined request for all niches, requesting:  
       "For each of these 20 niches, generate 3–5 primary keywords and 3–5 related long-tail keywords, including approximate search volume data, competition, or other high-level metrics if available."  
   • Secondary Skill (Optional): #224 – Oracle Ask A Question  
     - If more detailed or updated metrics are needed (e.g., to refine competition levels, gather real-time traffic data, etc.), the agent can optionally call #224 to query additional search tools.

3) REFINE & ORGANIZE DATA (#223)  
   - Once the keyword data is assembled, the agent calls #223 – Powerful LLM Prompt-to-Text Response.  
   - This step organizes/refines the data:  
     • Ensures each niche is clearly labeled.  
     • Groups primary and long-tail keywords.  
     • Supplements each keyword set with any final approximate metrics in a clean, consistent structure.

4) RETURN FINAL OUTPUT ([keyword-research])  
   - The refined, tabulated data for all 20 niches—each with 3–5 primary keywords and 3–5 long-tail keywords plus approximate search and competition metrics—constitutes the final subagent output.

F) SILOS (IF NEEDED)  
• SILO 1: Gather Niche & Requirements → Compose/Send Prompt (#218 or #224)  
• SILO 2: Refine & Format Data (#223) → Output [keyword-research]

SubAgent #2 - Diagram

Expand Diagram

A) SUBAGENT SUMMARY  
"Product Curator" searches for and compiles 3–5 high-profit ClickBank products (with gravity, average earnings, or other metrics) for each of the 20 ranked niches from Subagent 1.

B) FINAL TASK OUTPUT  
[product-recommendations] – a structured list, for each of the 20 niches, of 3–5 recommended ClickBank products along with any available metrics (e.g., gravity, average earnings, popularity, etc.).

C) SUBAGENT INPUT  
• {task-requirements} – the overall task instructions (includes the user’s requirement to find profitable ClickBank products for each niche).  
• [ranked-niche-list] – the output from Subagent 1 (the list of 20 ranked niches).

E) SUBAGENT TASK SUMMARY  
1) Use Skill #224 (Oracle Ask A Question) with a consolidated query, e.g. “For each of these 20 niches, list 3–5 top ClickBank products with the highest gravity score or earnings, along with available metrics (gravity, average earnings, popularity index, etc.).”  
2) Take the returned data and compile a final list of product recommendations per niche, noting the product name(s) and any relevant metrics.  
3) Output the result as [product-recommendations].

F) SILOS  
This subagent is small and self-contained, so no additional silos are needed. The task is performed in a straightforward single-query workflow:
• Input the 20 niches → Skill #224 (gather product details) → Compile and finalize [product-recommendations].

SubAgent #3 - Diagram

Expand Flow

A) SUBAGENT SUMMARY:  
“Final Aggregator & Reporter” compiles the niches, keywords, and products data from previous subagents into a single spreadsheet (CSV/XLSX) and also generates a short summary (DOCX/PDF) with recommendations.

B) FINAL TASK OUTPUT:  
[spreadsheet-and-summary-report] →  
• One spreadsheet file (CSV or XLSX) containing columns for the 20 niches, corresponding 3–5 keywords, 3–5 long-tail keywords, 3–5 products, and any associated metrics.  
• One text-based summary (provided as final text that can be saved as a .docx or .pdf), highlighting the top niches and giving recommendations.

C) SUBAGENT INPUT:  
• {task-requirements} (original instructions),  
• [ranked-niche-list] (from Subagent 1),  
• [keyword-research] (from Subagent 2),  
• [product-recommendations] (from Subagent 3).

E) SUBAGENT TASK SUMMARY (Step-by-Step Workflow):

1) Collect all available data:  
   - “Niche Researcher” output ([ranked-niche-list]) → top 20 niches ranked with brief profitability/competition notes.  
   - “Keyword Analyzer” output ([keyword-research]) → for each niche, 3–5 main keywords and 3–5 long-tail keywords (plus any metrics).  
   - “Product Curator” output ([product-recommendations]) → for each niche, 3–5 top ClickBank products with gravity/popularity/earnings info.

2) Generate spreadsheet (CSV or XLSX) with #223 (Powerful LLM Prompt-to-Text Response):  
   - Input: A carefully constructed message instructing the LLM to combine [ranked-niche-list], [keyword-research], and [product-recommendations], and format them as a comprehensive table.  
   - Output: Spreadsheet content in CSV or XLSX format containing columns for:  
       (a) Niche (rank/label)  
       (b) Primary Keywords (3–5)  
       (c) Long-tail Keywords (3–5)  
       (d) Product Names/IDs (3–5)  
       (e) Metrics (search volume, gravity, competition, earnings, popularity, etc.)

3) Generate short summary document with #223 (Powerful LLM Prompt-to-Text Response):  
   - Input: A separate instruction to the LLM, referencing the same combined data, to produce a concise textual overview of:  
       (a) The most promising niches (with reasons why)  
       (b) Key competition and ROI considerations  
       (c) Overall recommendations for next steps or marketing approaches  
   - Output: A text block that can be saved as a .docx or .pdf file

4) Return final deliverables:  
   - [spreadsheet-and-summary-report]  
     • The spreadsheet file (in .csv or .xlsx)  
     • The summary document (in .docx/.pdf or raw text suitable for converting)

F) SILOS (if helpful to visualize):
1) SILO 1: Gather Data  
   - Inputs: [ranked-niche-list], [keyword-research], [product-recommendations]  
   - Action: Internally collect/merge these into a combined data structure.

2) SILO 2: Create Spreadsheet  
   - Action: Use #223 → “Format into CSV or XLSX columns with all metrics and data.”  
   - Output: CSV/XLSX data.

3) SILO 3: Create Summary Text  
   - Action: Use #223 → “Generate short textual summary with recommendations.”  
   - Output: Text summary.

4) SILO 4: Deliver Final Files  
   - Combine the spreadsheet and the summary text into final deliverables.  
   - [spreadsheet-and-summary-report].