The Bootstrapped Founder · 2026-02-06 · 17 min
After six months of building Podscan almost exclusively with Claude Code, Arvid shares the configuration and prompting strategies that make agentic coding actually work. From connecting Claude to your browser with the - chrome flag so it can visually inspect your app, to the "Ralph Wiggum loop" that keeps the agent iterating until a task is truly done, to the permission settings that prevent it from nuking your database - these are the practical lessons that separate productive Claude Code users from those constantly cleaning up messes. Plus: why testing is Claude Code's superpower, and how to build a system prompt that turns raw code generation into genuine collaboration. This episode of The Bootstraped Founder is sponsored by Podscan.fm The blog post: The podcast episode: Check out Podscan, the Podcast database that transcribes every podcast episode out there minutes after it gets released: Send me a voicemail on Podline: You'll find my weekly article on my blog: Podcast: Newsletter: My book Zero to Sold: My book The Embedded Entrepreneur: My course Find Your Following: Here are a few tools I use. Using my affiliate links will support my work at no additional cost to you.
439: The Increasing Risk of Building in Public71 / 100438: AI Liability: The Landmines Under Your SaaS437: Data Is the Only Moat436: When Long-Term Investments Finally Pay Off434: Follow Your Passion (But Not Like That)