This document outlines the architectural shift from the current "Lootbox/RPC" model to the new "Code-Mode/UTCP" paradigm.
The LLM acts as a "Router". It decides on one tool, waits for the result, then decides the next step. This incurs a round-trip latency cost and token cost (re-reading history) for every single step.
Mnemonic shell functions for launching GitHub Copilot CLI with different models and modes.
| Prompt Mode | Interactive Mode | Model |
|---|---|---|
p |
pi |
Claude Opus 4.5 |
ps |
pis |
Claude Sonnet 4.5 |
Date: December 8, 2025 Goal: Reduce initial context consumption while maintaining full tool access and consistent Claude behavior
This project achieved a 54% reduction in initial context (7,584 → 3,434 tokens) while improving tool discovery and enforcement. The key insight: Claude doesn't need verbose documentation upfront—it needs triggers to know when to load detailed context.
This talk tells a story. Not a tutorial with examples, but a narrative with a protagonist, obstacles, and transformation. The audience will remember the character and identify with the journey.