Back to CatalogContext Engineering explicitly separates context into static (always loaded) and dynamic (on-demand), ensuring the agent receives dense, high-signal information only when necessary to lower API costs.
Agentic AI
SDLC
Context Engineering
Static vs Dynamic Context for efficient token economy.
Intent & Description
⚠️ Problem
Passing entire codebases into every prompt is financially unviable, dilutes the model’s signal, and leads to expensive token burn with poor results.
💡 Solution
Explicitly separate context into static (always loaded rules and memory) and dynamic (on-demand retrieved documents and skills). This ensures the agent receives dense, high-signal information only when necessary. Source: Osmani, A., Saboo, S., & Kartakis, S. (May 2026). The New SDLC With Vibe Coding. Google.
Real-world Use Case
- Scaling AI to navigate large, complex repositories.
- Managing agent memory across long-lived development sessions.
- Optimizing token economy for production-grade coding agents.
Source
📌 TL;DR
Never load the full repository into context; always use progressive disclosure to fetch only the relevant files or skills.
Advantages
- Significantly lowers ongoing API costs.
- Improves output quality by reducing prompt bloat.
Disadvantages
- Requires dedicated engineering effort to design retrieval mechanisms.
- Overkill for tiny, disposable prototype scripts.
- Not suitable for tasks that require no domain-specific knowledge or persistent state.