On Time Edge — Marketing Copy Generator v2.0

A FastAPI service for generating brand-safe marketing copy with built-in guardrails and Sugarman direct-response copywriting principles. Supports template-based generation and LLM-powered copy (via Ollama or OpenAI-compatible APIs).

What's New in v2.0 — Six Sugarman Enhancements

1. Trigger Detection — Scans generated copy for Sugarman's 30 psychological triggers (involvement, curiosity, urgency, specificity, etc.) and reports which are present, where, and how strongly.

2. First-Sentence Library — 20+ proven opener patterns across five concept types (curiosity, story, contrast, statistic, direct challenge) tuned for manufacturing audiences.

3. Auto-Objection Detection — Infers likely buyer objections based on the target audience (VP Ops, COO, Plant Manager, IT Director, Supply Chain Director) and checks whether the copy addresses them.

4. Concept Extraction — Identifies the core "big idea" in the copy, scores its strength (0-10), and suggests improvements if weak.

5. Interactive Refinement — POST /refine endpoint for iterative copy improvement with feedback-driven rewrites.

6. A/B Variant Generation — POST /variants generates 2-10 copy variants with different opener styles, then compares their trigger counts and slippery-slide scores.

Prerequisites

• Python >= 3.10

Setup

# Clone the repository
git clone https://github.com/dillonote/On-Time-Edge.git
cd On-Time-Edge

# Create and activate a virtual environment
python -m venv .venv
source .venv/bin/activate   # Linux / macOS
# .venv\Scripts\activate    # Windows

# Install dependencies
pip install -r requirements.txt

# Copy the example environment file and fill in your values
cp .env.example .env

Running locally

# Start the development server (auto-reload)
uvicorn main:app --reload

# Start on a specific port
uvicorn main:app --reload --port 8000

# Quick CLI test (template mode, no server needed)
python main.py

Testing

pytest

API Endpoints

Method	Path	Purpose
POST	/generate	Generate copy with full Sugarman analysis
POST	/refine	Iteratively refine existing copy
POST	/variants	Generate A/B test variants
GET	/triggers	List all 30 Sugarman triggers
GET	/openers	Browse first-sentence library
GET	/objections/{audience}	View objections for an audience
GET	/health	Health check

API Usage

POST /generate

Generate copy with full analysis:

curl -X POST http://localhost:8000/generate \
  -H "Content-Type: application/json" \
  -d '{
    "asset_type": "email_single",
    "offer_name": "On Time Edge Manufacturing Workshop",
    "target_audience": "VP of Operations",
    "primary_outcome": "measurable scheduling improvement in 12-18 months",
    "key_benefits": ["Reduce rework by 40%", "Make constraints visible"],
    "offer_details": "Book a 20-minute walkthrough.",
    "cta": "Schedule a Workshop",
    "provider": "template"
  }'

POST /refine

Iteratively improve copy:

curl -X POST http://localhost:8000/refine \
  -H "Content-Type: application/json" \
  -d '{
    "original_copy": "We help manufacturers schedule better.",
    "feedback": "Make it more urgent, add a specific dollar amount, open with a question",
    "asset_type": "email_single",
    "target_audience": "COO"
  }'

POST /variants

Generate A/B test variants:

curl -X POST http://localhost:8000/variants \
  -H "Content-Type: application/json" \
  -d '{
    "asset_type": "email_single",
    "offer_name": "On Time Edge APS Implementation",
    "target_audience": "Plant Manager",
    "primary_outcome": "a schedule that survives contact with reality",
    "key_benefits": ["Reduce rework", "Make constraints visible"],
    "offer_details": "Book a 20-minute walkthrough.",
    "num_variants": 3,
    "provider": "template"
  }'

Environment Variables

Variable	Purpose	Default
PROVIDER	Default LLM provider (template, ollama, compatible)	template
MODEL	Model name for LLM providers	llama3.1
TEMP	Temperature for LLM generation	0.6
OLLAMA_URL	Ollama API endpoint	http://localhost:11434/api/chat
COMPAT_BASE_URL	OpenAI-compatible API base URL	(none)
COMPAT_API_KEY	API key for compatible provider	(none)
BRAND_PROFILE_JSON	Path to custom brand profile JSON	(built-in defaults)

Project Structure

On-Time-Edge/
├── main.py             # FastAPI app + all 6 enhancement engines
├── test_main.py        # Test suite (pytest)
├── requirements.txt    # Python dependencies
├── .env.example        # Example environment variables
├── .gitignore          # Git ignore rules
├── CLAUDE.md           # AI-assisted development guidance
└── README.md           # This file

Contributing

1. Create a feature branch from main.
2. Keep changes small and focused.
3. Add or update tests when behavior changes.
4. Open a pull request with a clear description.