After 400+ production agents: a 14-layer design stack for agentic AI systems
MaryamMiradi · x · 2026-11-18
Drawing on 400+ production AI agents, the author lays out a 14-layer design stack via an end-to-end airport AI system: product + orchestration, agent boundaries + context, state + knowledge, action + authority, interaction + runtime, recovery + evaluation, and observability + learning. Key lessons: permissions enforced outside the agent, retries without duplicate side effects, and testing decisions/evidence/tool calls—not just final answers. Core thesis: a correct answer can hide an unreliable process.
More from coding & agent
- Cognition launches Devin Bots: one bot that leads a team of agents to ship work — round · 2026-11-18
- Shared applications, not shared chat: a new blueprint for collaborating with agents — thechrisperry · 2026-11-18
- Microsoft publishes playbook for enterprise deployment of governed AI agents on Windows via MXC — pavandavuluri · 2026-11-18
- AWS tests three OAuth patterns for agents calling Snowflake Cortex via MCP — AWS ML Blog · 2026-11-18
- A16z partner Anish Acharya curates a list of the sharpest thinkers on agent swarms — astrange · 2026-11-18
- Netlify CLI 27.18.0 moves auth tokens from plain text into OS keychains — jasonkneen · 2026-11-18