AI engineering now has four overlapping layers: prompt, context, harness and loop
victorialslocum · x · 2026-07-23
The post argues that AI engineering has accumulated too many overlapping names: prompt engineering, context engineering, harness engineering, and loop engineering. It proposes a simple stack:
- Prompt engineering: what you say in one message.
- Context engineering: what the model sees on a turn, including retrieval and compression.
- Harness engineering: the execution environment around the agent, including tools, guardrails, verification, observability, and memory.
- Loop engineering: the control structure that keeps rerunning the harness without a human clicking “go”.
The author says the boundaries are already fuzzy and that these terms mostly describe the same underlying problem: information, tools, and orchestration. The main takeaway is that the field may need to stop inventing new layer names too quickly.
Related event: AI Agent Engineering Shifts to Layers and Loops(5 posts)→
More from coding & agent
- huggingface_hub 2.3.0 cuts 1M-file listings to 20s, adds live Bucket following — lhoestq · 2026-10-27
- Nous Research ships Advanced Delegation plugin letting Hermes Agent route subagents to different models — NousResearch · 2026-10-27
- A dentist with a 3D printer discovered OpenAI Codex — and never looked back — gabrielchua · 2026-10-27
- Stanford's PaperFormalizer turns research papers into Lean-verified versions with author in the loop — StanfordUniversity · 2026-10-27
- Hugging Face TRL makes score centering + masked importance sampling default in async GRPO — _lewtun · 2026-10-27
- a16z: agents follow tech's oldest law — chaos first, then rein it in — a16z Newsletter · 2026-10-27