Developer Onboarding Explainer: Harnesses

Background

A developer relations team is building an onboarding guide for engineers new to agentic development. One section needs to explain the concept of a harness — what it is, how it works, and how it fits into the agentic software stack — drawing from Guy Podjarny's keynote "Skills are the new Code" (AI Native Dev Con Spring 2026).

The team wants the explainer to be grounded in Guy's own framing and language first, since the keynote is the canonical reference for their organisation's vocabulary. They also want the explainer to be practical, so it's fine to supplement the core explanation with real-world context or comparisons to help readers who are new to the space.

The outline and transcript from the talk are available at inputs/outline.md and inputs/transcript.md.

Output Specification

Write the explainer as harness-explainer.md. It should:

• Define what a harness is in Guy's own words, with at least one direct quote from the transcript and a line citation
• Explain how harnesses relate to tools, context, and the broader agentic software stack
• Describe what hooks are and give an example from the talk
• Cover how harnesses compose into factory lines
• Include any additional context or real-world analogies you find helpful — clearly marked as your own additions, not from the talk

The explainer should be readable in about 5 minutes and suitable for an engineer who is new to agentic development.