Introduction
In this post, I will list the key components I am building/using, explain what each does, and how they will eventually work together to streamline my low-friction output system.
Why Components Matter
- Easier visibility into what workflows can be streamlined and what is missing for certain goals.
- Building/Updating workflows becomes easier and more flexible once key components are robustly developed.
What Components I’m Identifying
Goal: have small, useful modules ready so I can combine them freely.
Input Modules
- Voice capture (ready)
- Screenshots (not started)
- Note capture (ready)
- Clip from web browser (ready, needs update and documentation)
Output Modules (termination display or storage)
- Emacs Org-mode
- Emacs Org-mode -> exported via ox-hugo as public blogs
- Apple Notes
- File system
Formatting Modules
- LLM structuring
- Emacs Smart Copy (Org -> Markdown -> Clipboard) and Smart Paste (Clipboard Markdown -> Org)
- Pandoc-based processing for compatibility
Intelligent Process Modules
- ChatGPT discuss
- gptel
- Trae with Claude for coding
- MCP agent
- LLM with tag to know better context of me
Retention/Schedule/Notification Modules
- Anki flashcards
- Cheat sheets
- System cron jobs
- ChatGPT scheduled tasks
- Emacs Org schedule (unsure if Emacs can support proactive notifications)
TTS/STT Module (special due to reading challenges and speaking practice needs)
- Speechify on web (desktop version is weak)
- Voice input
Glue Modules
- MCP agent
- gptel and other AI tools inside Emacs for batch processing, easier storage/refile, chat management
- Trae and Cursor
- Scripts (to automate capture, conversion, placement)
Module Refinement: Building Clear, Composable Units
Goal: Define clear, self-contained modules that can be flexibly combined to build workflows without starting from scratch.
Input Modules
Purpose: Bring raw data into the system.
- Voice capture (iPhone/Whisper) [ready]
- Screenshot capture (iPhone, Snagit, OBS) [not started]
- Note capture (Apple Notes, Emacs capture) [ready]
- Web clipping (Org Protocol, Readwise, Raindrop) [needs refinement]
Processing Modules
Purpose: Clean, format, or structure data for use.
- LLM structuring (ChatGPT deterministic prompts)
- Pandoc conversion
- Emacs Smart Copy/Paste
Intelligent Helper Modules
Purpose: Provide structured assistance, brainstorming, and idea refinement.
- ChatGPT body double
- gptel (inside Emacs, clarify context management)
- Trae/Claude for coding help
Storage/Output Modules
Purpose: Termination point for usable outputs.
- Emacs Org-mode
- Ox-hugo export to public blogs
- Apple Notes
- File system
Retention/Review Modules
Purpose: Ensure learning is retained and processed over time.
- Anki flashcards
- Cheat sheets
- Emacs Org agenda
- System cron jobs
- Email and ChatGPT scheduled tasks
Interaction Modules
Purpose: Assist reading, reflection, and practice.
- Speechify (TTS for reading challenges)
- Voice input for ideation and practice
Glue Modules
Purpose: Connect and automate modules for seamless workflow.
- MCP agent (clarify capabilities)
- gptel scripts
- Trae/Cursor
- Custom scripts for capture, conversion, and filing