Agentic Skill Decay

August 31, 2026

Mastery still comes from doing the reps.

Before agents, I got my reps as part of writing code: try different approaches out, debug what went wrong, review other’s code, read a lot. Agents can skip much of that work, so building your reps has to be deliberate.

If I was new to the industry, I’d try to form a hypothesis before prompting. Ask “why” a lot, read the diffs, try to predict what might fail. Occasionally try to work through the problem myself manually.

Slide titled Mastery still comes from the reps: they used to arrive on their own, now you have to go and get them. The top row shows how the reps used to arrive as a side effect of writing code: try an approach, watch it fail, trace the bug, read someone's fix, carry the lesson. The bottom row shows what an agent does with the same problem, going from the problem straight through to a working patch in minutes, with watch it fail, trace the bug and read someone's fix greyed out. So put the reps back on purpose: form a hypothesis before you prompt, ask why and keep asking it, read the diffs, predict what might still fail, and occasionally work it through by hand. Footer: I still use agents aggressively, I just don't want the issue to close while my mental model stays where it was.

In my experience, good agent work depends on two abilities:

  1. Deep expertise: you understand the problem domain well enough to define a good outcome. Understanding your user/product/business is part of this.
  2. Applied judgment: use your taste to turn this into a clear, testable plan by choosing the right context, constraints, tests and verification.