Pharos Oracle
Most people open an AI assistant and start typing. The ones who get the most out of it set up a few things first. Working with AI is a craft, and Pharos Oracle collects the practices that hold up, in plain English, in the order that matters.
Imagine bringing in a brilliant consultant to think through a problem with you. They are sharp, but they do not yet know your industry, your tools, or what 'good' looks like in your work. So you brief them. Every AI tool, from chat assistants to agentic workflows, walks in the same way. Pharos Oracle is here to close that gap.
Why this field guide exists
Most workplace AI today is one layer: the chat window. Pharos Oracle is the field guide to the layer beneath it.
Read the transcript
Most people at work today use AI through a chat window. Type a question, read an answer, open Teams Copilot, ask ChatGPT, close the tab. This is the layer everyone sees.
The Boston Consulting Group has a name for the gap behind it. The silicon ceiling. 75% of leaders use AI several times a week. Only 51% of frontline workers do.
Underneath the chat, there is a different paradigm. Same models, different surface. AI inside the editor where the work actually happens. It reads your codebase. It edits across files. It runs the terminal. It fixes its own mistakes. It builds with you, not beside you. This is the difference between asking AI a question and operating with AI as a collaborator.
Battery Ventures surveyed enterprise leaders in late 2025. 33% already run agentic AI in production. 48% planned to within 12 months. The interest is there. The deployment is not.
Pharos Oracle is the field guide that closes that gap. It teaches Visual Studio Code, GitHub Copilot, and the agentic workflows beneath them. The files an AI agent reads on every request. The rules that fire automatically. The skills that load only when the work calls for them. In plain English. In the order that matters. For people who have never written a line of code.
When the people in your organization understand the layer beneath the chat window, AI stops being a feature on a sidebar and starts behaving like a force multiplier they own. That is the gap this site closes.
The workspace matters as much as the prompt.
Most AI advice focuses on the prompt, the exact words you send the model. The workspace around the prompt matters just as much, and almost nobody teaches it. The files your AI reads on every request, the rules that fire when you save, the skills that load only when invoked. None of that is phrasing. It is architecture.
A well-engineered workspace lets a short, focused prompt do work that a bare environment could not match with a thousand-token essay. Same model, same question, completely different output. Pharos Oracle is the field guide for that architecture: the patterns, the layers, and the routing decisions that turn a wandering agent into a focused one.
Modules
Each one stands alone. Read them in any order. New modules land as the topics become worth teaching.