🤖 Inside the Agent Layer · Overview

Agent Layer at a Glance: The Infrastructure That Lets AI Agents Act

The agent layer of an AI factory on one page: the eight components between a model and real work (sandboxes, durable execution, tool gateways, identity, memory, observability and RL environments), their headline numbers, one agent task followed through every component, who builds what, a timeline, what goes wrong, and a builder for your own agent stack. Everything links to the eight-part series.

Explore the layer ↓ Start at Part 1 →
⚠️ About the numbers. Everything here comes from the eight articles (as of 28 October 2026), which cite their sources. Figures marked ⚠️ are reported or secondary; company figures are labelled as such; the builder matrix, risk labels, stack builder and quiz are editorial. Disclosure: Claude, made by Anthropic, helps write this site; Anthropic created MCP, and Claude Code and Anthropic features appear in the series alongside OpenAI, Microsoft, Google and others.

1. The eight parts

Tap a component.

2. One task, every component

A user asks a coding agent to fix a bug overnight. Press Play. A simplified composite.

Press Play.

3. Headline numbers

One or two figures per part. Tap to open the article.

4. Who builds what

Which kinds of company typically supply each component, as read across the series. ● major role · ◐ some · — little or none. Tap a cell.

Tap a cell.

5. Timeline, 2018 → 2026

Tap a milestone.

6. What can go wrong in each component

7. Build an agent stack

Four questions, one suggestion per component. An editorial starting point drawing on each part.

8. Quiz

Eight questions, one per part.

Answer all eight.

9. Reading paths

Related

Sibling hubs: Inside the ML Layer at a Glance, Inside the Data Layer at a Glance and Evaluation at a Glance. On the AI factory site plan, agent sandboxes and gateways sit in Data Hall B. See also agent runtimes in the ML Layer, sandboxing AI agents and production agent architecture.