Lauren Tan calls pstack her personal set of skills for careful engineering work. She says this set lets her reliably bring 2,000 pull requests into production each month.

Tan defines verification as an agent’s ability to check its own work and keep going until it succeeds. For this, she recommends giving the agent a small command-line utility with scripts for working with and debugging the app, rather than Markdown instructions alone.

Claim check:

  • Lauren Tan calls pstack her personal set of skills for careful engineering work. (confirmed by the publication itself: evidence; «In this series of posts, I’m going to show you how I use pstack , my personal set of skills for doing rigorous engineering work.»)
  • She says this set lets her reliably bring 2,000 pull requests into production each month. (confirmed by the publication itself: evidence; «It’s allowed me to ship 2,000 PRs a month to production with high confidence.»)
  • Tan defines verification as an agent’s ability to check its own work and keep going until it succeeds. (confirmed by the publication itself: evidence; «verification means that an agent can verify its own work. It can keep going until it succeeds at its task»)
  • For this, she recommends giving the agent a small command-line utility with scripts for working with and debugging the app, rather than Markdown instructions alone. (confirmed by the publication itself: evidence; «we prefer to give agents tools rather than just markdown. For verification skills, this means creating a small CLI that scripts interaction and debugging of your app in a small, agent friendly utility.»)

Publications:

score 65.8 out of 100 · kind: guide