vibeline/TODO.md

VibeLine TODO List

Steps to glory, aka make this thing awesome forever:

• Allow plugins to execute stuff using MCP Tools
• Convert transcribe step into a plugin
• Allow plugins to have an "input", i.e. which dir they should watch

After that:

• Create an "archive" plugin that moves old stuff to a separate folder
• Create even more plugins
• Dockerize it for easy deployment

Old vibe TODO list:

• Set up basic project structure
• Create voice memo processing system
• Implement dynamic plugin loading system
• Create blog post plugin
• Create app idea plugin
• Create summary plugin
• Set up proper directory structure for outputs
• Implement file organization system
• Add timestamp-based file naming
• Make scripts executable
• Extract llama models to ENV variables
• Create action_item plugin
• Use python library for pluralization
• Use more structured files for the plugins & put metadata into those structured files
• Write README.md with project overview, installation instructions, usage examples, plugin system documentation, and directory structure explanation
• Create LICENSE (MIT recommended)
• Create post-processing step
• Create graphic generation plugin (SVGs)
• Create image generation plugin (jpgs - using a local StableDiffusion model)
• Add error handling for Ollama connection issues
• Implement plugin validation system
• Add support for custom output formats
• Create plugin template system
• Add tests for core functionality
• Implement logging system
• Add configuration file support (.env)
• Create plugin documentation system
• Modernize CLI structure: - Replace bash scripts with a unified Python CLI tool - Implement subcommands: run, watch, process, transcribe - Use click/typer for better CLI experience - Support commands like: - vibeline run - vibeline watch - vibeline process transcript.txt - vibeline transcribe memo.m4a