Business Idea Refinement Agent

An AI-powered business idea analysis tool that provides comprehensive feedback through Herman Poppleberry, your friendly AI business plan review assistant.

Features

• Intelligent Analysis: Uses Gemini Pro 2.5 to analyze business ideas with venture capitalist-level insights
• Audio Feedback: Generates personalized audio feedback using Gemini TTS with Herman Poppleberry's voice
• Email Delivery: Automatically sends analysis and audio feedback via email using Resend API
• Multiple Voice Styles: Choose from upbeat, serious, or flippant delivery styles

Example Run

Input: Baby QR Onesies Business Idea

The agent processed a real business idea about QR code-enabled baby onesies for Home Assistant integration.

AI-Generated Output (Herman Poppleberry's Analysis)

Organized Output Folder: agent/feedback/20250828_170126_baby-qr-onesies/

The agent automatically:

• Generated AI-suggested filename: baby-qr-onesies
• Created timestamped folder with descriptive name
• Produced comprehensive venture capitalist-style analysis
• Generated multiple formats for different use cases

Screenshots

Agent delivers analysis by email

Quick Start

1. Setup Environment

# Install dependencies
pip install -r requirements.txt

# Copy and configure environment variables
cp .env.example .env
# Edit .env with your API keys

2. Configure API Keys

Add your API keys to .env:

GEMINI_API_KEY="your_gemini_api_key_here"
RESEND_API_KEY="your_resend_api_key_here"  # Optional for email delivery
USER_EMAIL="your_email@example.com"

3. Run the Agent

# Auto-process all pending ideas
python business_idea_agent.py

# Choose voice style and skip email
python business_idea_agent.py --voice-style serious --no-email

Usage Examples

Basic Auto-Processing

# Processes all .md files in agent/user-ideas/pending/
python business_idea_agent.py

Custom Voice Style

python business_idea_agent.py --voice-style serious

Skip Email Notifications

python business_idea_agent.py --no-email

Voice Styles

• upbeat (default): Encouraging and enthusiastic tone
• serious: Authoritative, judge-like delivery
• flippant: Sardonic, defeatist tone

Output

The agent generates:

• Text Analysis: Saved to agent-feedback/analysis_TIMESTAMP.txt
• Audio Feedback: Saved to agent-feedback/business_analysis_TIMESTAMP.wav
• Email: Sent with both text summary and audio attachment (if configured)

Workflow

1. Agent Review: Gemini Pro 2.5 analyzes your business idea using the carefully crafted system prompt
2. TTS Generation: Gemini 2.5 Flash Preview TTS creates audio feedback with Herman Poppleberry's voice
3. Email Delivery: Resend API sends results with audio attachment and summary

Models Used

• Text Analysis: Gemini 2.0 Flash Exp
• TTS: Gemini 2.5 Flash Preview TTS (Voice: Charon)

File Structure

├── business_idea_agent.py      # Main agent implementation
├── design/
│   ├── system-prompt.md        # Herman Poppleberry's instructions
│   └── gemini-audio-instruction.md  # TTS voice guidelines
├── agent/
│   ├── feedback/              # Generated outputs
│   └── user-ideas/
│       ├── drafting/          # Ideas being developed
│       ├── pending/           # Ideas awaiting analysis
│       └── evaluated/         # Processed ideas
└── requirements.txt           # Python dependencies

Workflow

1. Draft: Create business ideas in agent/user-ideas/drafting/
2. Queue: Move ready ideas to agent/user-ideas/pending/
3. Process: Run python business_idea_agent.py to auto-process all pending ideas
4. Review: Check results in agent/feedback/ and moved files in agent/user-ideas/evaluated/