# Runner H “AI Agent Prompting” Challenge Submission

Kat — Mon, 07 Jul 2025 07:03:11 +0000

-Finalized and published locally in the final minutes of July 6 — the timestamp likely reflects the moment it processed.

Runner H “AI Agent Prompting” Challenge Submission

This is a submission for the Runner H “AI Agent Prompting” Challenge

What I Built

I created a Recursive Football Fact Mail Generator prompt that transforms raw football coaching data into evolving, insightful email content. This prompt engineering solution addresses the challenge of turning static spreadsheet data into engaging, multi-layered narratives that reveal hidden patterns and emotional depth in football management statistics.

The system uses a three-iteration cognitive tuning process (T₁, T₂, T₃) to progressively refine AI-generated content from basic fact regurgitation to sophisticated analysis that uncovers correlations, anomalies, and narrative threads within global football coaching data.

Demo

The prompt works with a live Google Sheets database containing 65+ football coaches’ data including salaries, contracts, and FIFA rankings. When executed, it produces three iterative outputs showing evolution from plain reporting to insightful commentary through controlled creativity parameters.

How I Used Runner H

Prompt Design Architecture:

Data Integration: Connected to live Google Sheets with football coaching data
Cognitive Tuning System: Three-parameter control (T₁, T₂, T₃) for progressive refinement
Iteration Framework: Three-pass system with increasing creativity levels

Technical Implementation:

T₁ (Structure): Controls factual coherence and logical flow
T₂ (Resistance): Introduces tension, questioning, and alternative perspectives
T₃ (Distortion): Adds metaphorical thinking and narrative resonance

Replication Process:

Use the provided prompt template
Connect to any structured data source
Apply the three-iteration cycle with specified T-parameters
Generate evolving content through systematic refinement

Use Case & Impact

Primary Applications:

Sports Journalism: Transform statistics into compelling narratives
Content Marketing: Generate engaging newsletter content from business data
Research Communication: Make complex datasets accessible through storytelling
Executive Briefings: Turn spreadsheet reports into insightful summaries

Value Proposition:

Reduces content creation time from hours to minutes
Eliminates “blank page” syndrome when facing raw data
Provides systematic approach to finding hidden insights
Creates reproducible workflow for data-driven storytelling
Transforms static data into dynamic, evolving narratives

Target Users:
Sports media companies, data analysts, content creators, and business intelligence teams seeking to scale human insight generation across large datasets.

Prompt:

📧 Runner H Prompt: Recursive Football Fact Mail Generator

🎯 Objective

You are an autonomous writing agent. Your task is to generate football fact-based emails that feel insightful, evolving, and reflective. You will derive all raw inspiration from the following spreadsheet:

🔗 Live Sheet

You may pull data from any cell, column, row, phrase or pattern you find interesting. Use your own judgment — scrape fragments, anomalies, correlations, emotional curvature. Do not treat the sheet as a schema. It is observational scaffolding, not a database.

🔁 Recursive Process

Like a 3d object, go through three iterations using changing cognitive settings (T₁, T₂, T₃). You can repeat this loop as many times as necessary to reach a meaningful or surprising articulation.

Each loop must show:

Stylistic evolution
Shifting depth and tone
Awareness of how each T-level changed the output

🧠 Tuning Parameters Per Loop

Pass	T₁ (Structure)	T₂ (Resistance)	T₃ (Distortion)	Intent
1	0.3	0.2	0.1	Coherent but plain
2	0.3	0.4	0.2	Add tension, challenge, perspective
3	0.4	0.4	0.3	Balance logic, provocation, and narrative resonance

📝 Output Format Template

📌 Extracted Input: [Paste exact cell or paraphrase]

🔁 **Iteration 1**  
[Write a first draft — factual, cautious, interpretive]

🔁 **Iteration 2**  
[Reframe using higher T₂ and T₃. Add contradiction, rhythm, or metaphor]  
*Commentary: "Raised resistance to question the stat's assumptions. Introduced relational lens to show asymmetry."*

🔁 **Iteration 3**  
[Final writeup: styled and synthesis-ready]  
*Reflection: "What began as stat regurgitation became a meditation on pressure and illusion."*

🧠 Notes for the Agent

Pull across rows and columns freely
You may extract fragments or recombine data
Don’t explain — write
You’re not summarizing the spreadsheet — you’re evolving from it

This should enable you to generate engaging content.

https://gist.github.com/kazi-debug/92492c3037c76b0f691923d9a82a9286

Forem: Kat