AI-Powered Project Checklist and Assessment System

Overview

This project provides an AI-driven system to help define, verify, and assess project readiness through a structured checklist approach. It uses AI models (Ollama and Google's Gemini) to generate checklist criteria, validate them interactively, simulate a debate between AI personas, and produce a final recommendation.

Features

• Generate project readiness checklists tailored to your industry.
• Interactive multiprompt verification of checklist items.
• AI-simulated debate between optimistic and critical personas.
• Final AI-generated recommendation to proceed or reconsider.
• Saves detailed assessment results to project_assessment_output.json.
• Supports loading predefined input from project_assessment_input.json for automated assessments.

Requirements

• Python 3.x
• requests library
• google.generativeai Python client
• Environment variable GEMINI_API_KEY set with your Gemini API key.
• Optional: Ollama server running locally at http://localhost:11434 for enhanced performance.

Usage

1. Set the GEMINI_API_KEY environment variable.
2. (Optional) Start the Ollama server locally.
3. Run the script:

   python main.py

4. Follow the interactive prompts to enter your project industry and answer validation questions.
5. Review the AI-generated analysis, debate, and final recommendation.
6. Check project_assessment_output.json for full output data.

Logging

The script uses Python's logging module to provide runtime information and error messages.

License

This project is provided as-is for prototyping and educational purposes.

Project Diagram

graph TD
    A[Start] --> B{Define Checklist Criteria}
    B --> C{AI Generates Checkpoints}
    C -- Malformed/Error --> D[Clean/Default Checklist]
    C -- Valid --> E[Checklist Criteria]
    D --> E
    E --> F{Multiprompt Verification}
    F --> G{AI Generates Questions}
    G --> H[User Answers Questions]
    H --> I{AI Analyzes Answers}
    I --> J[Analysis Report]
    J --> K{Counterargument Simulation}
    K --> L{AI Generates Pro-Argument Code}
    L --> M[Execute Pro-Argument Code]
    M --> N[Pro-Argument]
    K --> O{AI Generates Counter-Argument}
    O --> P[Counter-Argument]
    N & P --> Q{AI-Generated Conclusion}
    Q --> R[Final Recommendation]
    R --> S[End]

Loading