Browse Teach Curate

Trusting the Output of an Unsupervised Agent Run

Build the evidence trail — tests, diffs, logs — that lets you accept work you did not watch happen, and decide what level of proof each task requires.

Sonny Text

Trusting the Output of an Unsupervised Agent Run — Claude Code in Action

Build the evidence trail — tests, diffs, logs — that lets you accept work you did not watch happen, and decide what level of proof each task requires. Taught in the course “Claude Code in Action” (ht…