AI Core Skills
Section

AI Core Skills

One in-demand AI skill at a time: what top roles pay for it, where it came from, how to do it with AI, how to check your work, and a kit you can use tonight.

7 articles52 min total

Start reading →

#1 Agent evals

  1. Agent evals: the short versionWhat top PM roles pay for agent evals, what an eval actually is, and the kit you can use tonight.2 min read
  2. What top PM roles pay for agent evalsThe real job postings, checked, and an honest read on what the numbers mean.3 min read
  3. Where agent evals came from, and what they meanFrom 1930s holdout tests to agents that cheat on benchmarks: why evals exist and why it's a PM skill.21 min read
  4. How to build an agent eval with AIWhat AI does and what you do, step by step, with real task files, graders and a judge prompt.6 min read
  5. How to validate your agent evalEight gates that prove your eval measures what you think it measures, with real results.5 min read
  6. The agent evals kitThe Claude Code skill, two agents, ten prompts, the repos and seven practice builds.10 min read
  7. Agent evals reference: real tasks, tools, learning pathReal benchmark task files, the October 2026 tool landscape, a learning path, the interview angle, and sources.5 min read