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’s job is to take all of the user’s ad campaign requirements (budgets, targeting, schedule, and so on) and produce a single JSON config file (“campaign_config.json”) that clearly sets out how the campaign will run.

B) FINAL TASK OUTPUT  
• A structured file named “campaign_config.json” containing all final ad campaign settings (budget, targeting, schedule, etc.).

C) SUBAGENT INPUT  
• Desired budget parameters / constraints (including any recommended or “optimal” budget).  
• Demographic and interest targeting (age range, location, gender, interests).  
• Schedule/duration details (start date + 7 days, with peak days Thu–Sun).  
• Other relevant instructions or overrides for the ad campaign.

E) SUBAGENT TASK SUMMARY (Step-by-Step Skill Flow)  
1. (Optional) Skill #223 – “Powerful LLM Prompt-to-Text Response”  
   – Purpose: If the user needs a recommended budget or any extra strategic insight (e.g., to propose an “optimal” budget) before generating the final config, the subagent can consult this skill.  
   – Input:  
       • Prompt restating or clarifying the user’s performance/conversion goals.  
       • Query about recommended daily or total budget based on 7‑day run and a desired 10% conversion.  
   – Output:  
       • Text response (containing recommended total budget or daily budget distribution, if needed).

2. Skill #190 – “Write or rewrite text based on instructions”  
   – Purpose: Generate the final JSON configuration file using all campaign parameters (including the recommended budget if that was retrieved in the previous step).  
   – Input:  
       • Instructions to create a valid JSON document named “campaign_config.json.”  
       • Include top-level keys such as “campaignName,” “budget,” “schedule,” “locationTargets,” “demographics,” “interests,” etc.  
       • Incorporate the validated or “recommended” budget figure.  
   – Output:  
       • The final text of “campaign_config.json” containing the fully structured ad campaign settings.

F) SILOS  
• SILO 1: (Optional) BUDGET RECOMMENDATION  
   – Use Skill #223 to get recommended budget (if needed).  
• SILO 2: GENERATE JSON CONFIG  
   – Use Skill #190 to compile all inputs (including final budget, demographic, scheduling, interests) into “campaign_config.json.”

SubAgent #1 - Diagram

Expand Diagram

A) SUBAGENT SUMMARY  
This subagent (TrackingSetupAgent) is responsible for preparing easy‑to‑use “tracking details” and “pixel code” so we can track inbound clicks and conversions for the TikTok Mastery campaign, specifically by generating a “tracking_details.txt” (containing explanations, sample UTM parameters, and instructions on where to place them) and “tiktok_mastery_pixel.js” (a JavaScript snippet to track key events).

B) FINAL TASK OUTPUT  
Two files:  
• tracking_details.txt — Explanation of UTM usage (e.g., ?utm_campaign=tiktok_mastery) and notes on snippet placement  
• tiktok_mastery_pixel.js — A sample JavaScript code snippet for traffic/conversion tracking

C) SUBAGENT INPUT  
• Required details about how UTM tags should be structured (e.g., source, medium, campaign for “tiktok_mastery”)  
• Guidance on what the pixel snippet should track (e.g., clicks, conversions)  
• Any naming conventions or additional parameters needed

E) SUBAGENT TASK SUMMARY  
1) Receive instructions on the UTM parameters (source, medium, campaign) and the JavaScript snippet requirements.  
2) Use skill #190 (Write or rewrite text based on instructions) to generate “tracking_details.txt,” which explains how to implement UTM parameters (including standard parameters: utm_source, utm_medium, utm_campaign, etc.), plus any notes on placing them.  
3) Use skill #190 (Write or rewrite text based on instructions) again to generate “tiktok_mastery_pixel.js,” containing a sample JavaScript snippet to track clicks and conversions (e.g., using a button event listener or relevant code for the cart/checkout page).  
4) Output these two text-based files: “tracking_details.txt” and “tiktok_mastery_pixel.js.”

F) SILOS

Silo 1: Produce “tracking_details.txt”  
   • Input: Subagent instructions on UTM parameters.  
   • Action: skill #190 to write the textual explanation on UTM usage and snippet placement.  
   • Output: “tracking_details.txt”

Silo 2: Produce “tiktok_mastery_pixel.js”  
   • Input: Subagent instructions on needed JavaScript tracking snippet.  
   • Action: skill #190 to write the JavaScript file that handles click/conversion tracking logic.  
   • Output: “tiktok_mastery_pixel.js”

These two outputs complete the subagent’s workflow for producing the UTM and tracking code setup.

SubAgent #2 - Diagram

Expand Diagram

A) SUBAGENT SUMMARY  
This DataCollectorAgent will, each day for 7 days, fetch the ad performance metrics (clicks, cost, conversions, and geographic data) from an external ad platform and compile them into a single JSON file.

B) FINAL TASK OUTPUT  
A file named “daily_metrics.json” that contains, per day and per region, the date, number of clicks, number of conversions, cost, and any other relevant data needed for the campaign.

C) SUBAGENT INPUT  
1) “campaign_config.json” (which includes campaign ID, targeting details, and any external API credentials).  
2) Any required authentication tokens or API endpoints for retrieving daily ad performance data.

E) SUBAGENT TASK SUMMARY  
Below is the sequence of steps that this subagent will perform. Because there is no direct “fetch external ad metrics” skill listed, the subagent will rely on an external manual/API-integration step to actually retrieve the data and then transform it into the final JSON structure using an available text-generation skill (#190).

1) Receive Inputs:  
   - “campaign_config.json” (including campaign ID and targeting parameters).  
   - External credentials or instructions on how to query the ad platform’s API.

2) External Ad Platform Retrieval (Manual/API Integration):  
   - (Manual or Custom Code) Connect to the ad platform’s API or dashboard.  
   - Pull daily metrics (date, region, clicks, conversions, cost).  
   - Store this raw data in-memory or as a temporary text-based data structure.

3) Convert Metrics to JSON (#190 - Write or rewrite text based on instructions):  
   - Input: The raw data (step 2) plus instructions to produce structured JSON.  
   - Output: “daily_metrics.json” in a format containing:  
       [
         {
           "date": "YYYY-MM-DD",
           "region": "UK / USA / Canada",
           "clicks": <number>,
           "conversions": <number>,
           "cost": <currencyValue>,
           "conversionRate": <percentageOrDecimalIfDesired>
         },
         ...
       ]

4) Save/Output the “daily_metrics.json”:  
   - This file is updated each day of the 7-day campaign.  
   - By the end of the 7 days, it contains each day’s metrics from the relevant regions.

F) SILOS  
Silo 1: Data Retrieval  
• Manual/API step to query ad platform for daily metadata.

Silo 2: JSON Creation/Update  
• Use skill #190 to generate or update “daily_metrics.json” with properly formatted data.

––––––––––––––––––––––––––––––

That completes the DataCollectorAgent flow. The subagent’s final output to the larger workflow is “daily_metrics.json,” containing daily clicks, conversions, cost, region, and date over the 7‑day campaign.

SubAgent #3 - Diagram

Expand Flow

A) SUBAGENT SUMMARY:  
The CSVGeneratorAgent compiles the daily traffic data from the collected JSON metrics, calculates conversion rates, and outputs a finalized “TrafficSummary.csv” file (with columns for Date, Country, Clicks, Conversions, Cost, and Conversion Rate).

B) FINAL TASK OUTPUT:  
A single CSV file named "TrafficSummary.csv," containing one row per day (per region) that includes Date, Country, Clicks, Conversions, Cost (if available), and Conversion Rate (%).

C) SUBAGENT INPUT:  
• The “daily_metrics.json” file (which holds day-by-day data with date, clicks, conversions, cost, and region).  
• Any additional instructions needed for formatting the CSV (e.g., how to round numbers, how to handle missing data).

E) SUBAGENT TASK SUMMARY:  
1. Receive “daily_metrics.json” plus any formatting instructions as input.  
2. Parse the JSON data (date, country, clicks, conversions, cost) by using the LLM ability to handle text and JSON.  
3. Calculate the “Conversion Rate” by dividing “Conversions” by “Clicks” (e.g., conversions ÷ clicks).  
4. Construct the CSV text using the columns: Date, Country, Clicks, Conversions, Cost, Conversion Rate.  
5. Generate the CSV file ("TrafficSummary.csv") as a text output.

└─ Implementation Outline (Proposed Skills Flow)  
      a) Use #190 - “Write or rewrite text based on instructions”  
         – Prompt the skill to parse the daily_metrics.json input (as text) and transform it into properly formatted CSV rows, calculating conversion rate as needed.  
      b) Return final CSV text as “TrafficSummary.csv.”

F) SILOS:  
• SILO 1: Parsing & Calculation  
   – Parse the JSON data for each date/region.  
   – Calculate conversion rates.

• SILO 2: CSV Generation  
   – Convert parsed data into CSV-formatted rows.  
   – Finalize “TrafficSummary.csv” as the single output.

4 Template & Links

Expand Flow

A) SUBAGENT SUMMARY  
This subagent (“PerformanceAnalysisAgent”) takes the final Traffic Summary CSV from Subagent 4, analyzes it against the target 10% conversion goal, and returns two outputs:  
1) A short text report summarizing performance (e.g., how many total clicks vs. conversions, how close to 10% overall, any notable observations).  
2) A simple chart/graph (jpeg) visualizing clicks and conversions by day and region.

B) FINAL TASK OUTPUT  
• “analysis_report.txt” containing a concise text summary of overall performance and any noteworthy insights.  
• “performance_graph.jpg” visualizing the data from the CSV.

C) SUBAGENT INPUT  
• “TrafficSummary.csv” generated by Subagent 4.

E) SUBAGENT TASK SUMMARY (Step-by-Step)  
1) Subagent receives “TrafficSummary.csv.”  
2) Parse CSV Data (internally). This step is conceptual; no direct skill call is needed to open CSV, but we will feed the extracted data into subsequent skills.  
3) Summarize performance vs. 10% conversion goal (Skill #185 - Write or rewrite text based on instructions).  
   – INPUT: The CSV data (as text) plus prompt instructions to compose an analytic summary.  
   – OUTPUT: “analysis_report.txt” (summarized performance, whether 10% conversion was met, regional breakdown, etc.).  
4) Create a performance graph (Skill #227 - Return Graph Image Visualising Data).  
   – INPUT: Structured data (parsed from CSV) plus instructions to generate a simple line/bar chart with daily clicks vs. conversions by region.  
   – OUTPUT: “performance_graph.jpg.”

5) Subagent final output:  
   – “analysis_report.txt” (the text summary)  
   – “performance_graph.jpg” (the chart image).

F) SILOS  
• SILO 1: Data Parsing & Analysis  
   1. Input “TrafficSummary.csv.”  
   2. Internally interpret CSV data.  
   3. Summarize findings vs. 10% goal (#185). Output “analysis_report.txt.”

• SILO 2: Data Visualization  
   1. Pass the structured data from the CSV to #227.  
   2. Generate “performance_graph.jpg” for daily clicks/conversions.

5 Template & Links

Expand Flow

A) SUBAGENT SUMMARY  
This subagent (“PerformanceInsightsAgent”) takes the final “TrafficSummary.csv” from the other subagents, analyzes its contents (daily clicks, conversions, costs, and geographic breakdown), then generates a concise text-based performance summary and an optional visualization image that can be shared with stakeholders.

B) FINAL TASK OUTPUT  
• A text file named “PerformanceInsights.txt” summarizing the results (e.g., average daily clicks, total conversions, and whether we met the 10% goal).  
• An optional visualization named “PerformanceGraph.jpg” (bar/line chart) showing daily clicks vs. conversions by region, if desired.

C) SUBAGENT INPUT  
• “TrafficSummary.csv” containing all daily campaign data, broken down by date, region, clicks, conversions, cost, and conversion rate.  
• Any additional context or instructions about the specific metrics to highlight or analyze.

E) SUBAGENT TASK SUMMARY

1) Receive “TrafficSummary.csv”  
   - The subagent is provided with the final CSV generated from the previous workflow steps.

2) Parse CSV Content into Subagent Memory  
   - The subagent reads the CSV data (no specific skill needed, as reading it is an internal step).

3) Summarize Key Performance Indicators (Text)  
   - Using #190 (Write or rewrite text based on instructions) or #223 (Powerful LLM Prompt-to-Text Response), the subagent extracts pertinent metrics (daily average clicks, total cost, total conversions, overall conversion rate, top-performing country, etc.) and creates a text summary.  
   - OUTPUT: “PerformanceInsights.txt” (summarizing how well the campaign performed compared to the 10% conversion goal, plus any other meaningful insights).

4) [Optional] Generate Visual Chart of Results (Image)  
   - If desired, the subagent uses #227 (Return Graph Image Visualising Data) to create a simple bar or line chart from the daily data.  
   - OUTPUT: “PerformanceGraph.jpg”

5) Subagent Outputs Performance Files  
   - Provide the two final files as the subagent’s completed result.

F) SILOS

• SILO 1: DATA RECEPTION  
   - Input: “TrafficSummary.csv”  
   - Action: Internally read CSV contents for further analysis.

• SILO 2: TEXT SUMMARY GENERATION  
   - Use skill #190 or #223 to analyze CSV data, write a bullet-point or paragraph summary of key findings.  
   - Output: “PerformanceInsights.txt”

• SILO 3: VISUALIZATION (Optional)  
   - Use skill #227, providing a short mention of the data (dates, regions, clicks, conversions) to generate a meaningful image.  
   - Output: “PerformanceGraph.jpg”

• SILO 4: FINAL OUTPUT DELIVERY  
   - Provide “PerformanceInsights.txt” and optional “PerformanceGraph.jpg” for end users/stakeholders to review.