AI DevOps Podcast Richard Lander: DotNet Inspect - Episode 392
13 snips
Mar 9, 2026 Richard Lander, Principal Program Manager on the .NET team at Microsoft with decades shaping runtime and tooling. He traces .NET’s evolution to native AOT and explores inspect-style tooling inspired by Docker/kubectl. He unveils .NET Inspect, its markdown-first outputs, API diffing, LLM-friendly formats, SourceLink-driven source retrieval, and how CLI-first, pipeable tools enable AI-enabled developer workflows.
AI Snips
Chapters
Transcript
Episode notes
Three Layer View Of Dev Tooling
- Lander frames developer tooling as three durable layers: models, agents (CLIs/agents like Copilot CLI), and the dev SDK (language tooling and package managers).
- He believes investing in the CLI layer offers high leverage for durable developer experiences amid AI changes.
Prototype CLI Features Before Changing The Main CLI
- Do invest in the .NET CLI as a durable leverage point for integrating AI workflows rather than only building LLM-hosted services.
- Lander built a lightweight prototype (.NET Inspect) to prove CLI-centered ideas before changing the main CLI.
Docker Inspect Inspired Query First CLI Design
- Docker inspect and kubectl describe influenced .NET Inspect: command-line, query-first, JSON/structured output that agents and humans can consume.
- Lander emphasized power-user visibility into artifacts (images/packages) and queryability as core UX lessons.