Agent workflows and addiction to prompting

Brought to You By:

• Statsig — ⁠ The unified platform for flags, analytics, experiments, and more.

• Sonar – The makers of SonarQube, the industry standard for automated code review

• WorkOS – Everything you need to make your app enterprise ready.

—

Peter Steinberger ships more code than I’ve seen a single person do: in January, he was at more than 6,600 commits alone. As he puts it: “From the commits, it might appear like it's a company. But it’s not. This is one dude sitting at home having fun."

How does he do it?

Peter Steinberger is the creator of Clawdbot (as of yesterday: renamed to Moltbot) and founder of PSPDFKit. Moltbot – a work-in-progress AI agent that shows what the future of Siri could be like – is currently the hottest AI project in the tech industry, with more searches on Google than Claude Code or Codex. I sat down with Peter in London to talk about what building software looks like when you go all-in with AI tools like Claude and Codex.

Peter’s background is fascinating. He built and scaled PSPDFKit into a global developer tools business. Then, after a three-year break, he returned to building. This time, LLMs and AI agents sit at the center of his workflow. We discuss what changes when one person can operate like a team and why closing the loop between code, tests, and feedback becomes a prerequisite for working effectively with AI.

We also go into how engineering judgment shifts with AI, how testing and planning evolve when agents are involved, and which skills and habits are needed to work effectively. This is a grounded conversation about real workflows and real tradeoffs, and about designing systems that can test and improve themselves.

—

Timestamps

(00:00) Intro

(01:07) How Peter got into tech 

(08:27) PSPDFKit

(19:14) PSPDFKit’s tech stack and culture

(22:33) Enterprise pricing

(29:42) Burnout 

(34:54) Peter finding his spark again

(43:02) Peter’s workflow 

(49:10) Managing agents 

(54:08) Agentic engineering

(59:01) Testing and debugging 

(1:03:49) Why devs struggle with LLM coding

(1:07:20) How PSPDFkit would look if built today 

(1:11:10) How planning has changed with AI 

(1:21:14) Building Clawdbot (now: Moltbot)

(1:34:22) AI’s impact on large companies

(1:38:38) “I don’t care about CI”

(1:40:01) Peter’s process for new features 

(1:44:48) Advice for new grads

(1:50:18) Rapid fire round

—

The Pragmatic Engineer deepdives relevant for this episode:

• Inside a five-year-old startup’s rapid AI makeover

• When AI writes almost all code, what happens to software engineering?

• Why it’s so dramatic that “writing code by hand is dead”

• AI Engineering in the real world

• The AI Engineering stack

—

Production and marketing by ⁠⁠⁠⁠⁠⁠⁠⁠https://penname.co/⁠⁠⁠⁠⁠⁠⁠⁠. For inquiries about sponsoring the podcast, email podcast@pragmaticengineer.com.