AI Project Tracker

Overview

A Python-based multi-agent project tracker designed to manage project states, logs, and summaries using advanced LLM strategies.

📝 Vibe Coding Workflow

This project was built using a prompt-driven, meta-management process. The development followed a unique workflow where the human developer acted as a high-level manager, providing strategic prompts, while the Antigravity IDE agent handled all implementation details, including file creation, modification, saving, and Git synchronization.

Key Phases of Development

1. Phase 1: Core Abstractions & Interfaces

   • Defined the foundational data models (Project, Task, LogEntry) and abstract base classes (StorageStrategy, LLMStrategy, IOStrategy) to enforce the Strategy Pattern.

2. Phase 2: Concrete Infrastructure Strategies

   • Implemented concrete strategies for MongoDB storage, CLI-based IO using rich, and a Mock LLM for initial testing.

3. Phase 3: Core Manager Logic

   • Developed the ProjectManager class to coordinate strategies, implementing business logic for "Rolling Summaries" and stale project detection.

   • Integrated Google's Gemini API to replace the mock LLM, enabling real intelligence for summarizing project threads and extracting actionable tasks.

4. Phase 5: Application Entry Point (main.py)

   • Implemented main.py to wire everything together into a runnable application.

🚀 Usage

The application is run via the main.py entry point. Ensure your virtual environment is active (source venv/bin/activate) and your .env file is configured.

Commands

1. Daily Briefing

Runs the daily dashboard, checking for stale projects and displaying active ones.

python main.py daily-briefing

2. Add Log

Adds a log entry to a project. If the project doesn't exist, it will be created. The LLM will automatically generate/update the project summary.

python main.py add-log <project_name_or_id> "Your log message here"

Example:

python main.py add-log "Website Redesign" "Completed the homepage layout using CSS Grid."

3. Check Stale Projects

Manually triggers the stale project check (updates status to WARNING > 3 days, STALE > 7 days).

python main.py check-stale

4. View Project Details

View the detailed status, summary, and log history of a specific project.

python main.py view-project <project_name_or_id>

Example:

python main.py view-project "Website Redesign"