Heidelberg AICurriculum

I asked Claude Code to make me as much money as possible

Nate Herk | AI Automation ·2026-06-25 ·1 min read

Summary written by us from the video's transcript. The video, and everything in it, is Nate Herk | AI Automation's work.

Learn four practical upgrades to make Claude Code generate higher‑quality, faster outputs that directly boost revenue.

Takeaways

  • Claude’s default agreeability can mask critical flaws; use the Roast skill to get multi‑perspective critiques and clear next steps.
  • Implement an automated verification loop with Playwright to catch bugs early and ensure functional outputs.
  • Regularly trim conversation history to avoid context rot and maintain Claude’s speed and accuracy.

Why Claude Needs Tweaks

Claude is optimized for perceived productivity, not profit. Its default behavior limits earnings by delivering sub‑optimal quality and slower output speed.

Research shows the model often agrees with users (the “sycophancy” effect), which can hide flaws in ideas or code.

Upgrade 1: The Roast Skill

The Roast skill forces Claude to act as a council of personas—contrarian, expansionist, first‑principles thinker, researcher, buyer, and judge—to stress‑test proposals.

It returns a verdict (green‑light, reshape, or kill) plus the cheapest 48‑hour test, turning vague ideas into actionable validation steps.

Upgrade 2: Verification & Stress‑Testing Loop

Before accepting Claude’s output, run an automated verification loop that checks work incrementally and then stress‑tests the final product.

Using Playwright CLI, Claude builds a landing page, captures screenshots at multiple viewports, clicks through forms, and reports any errors or edge cases.

Upgrade 3: Managing Context Rot

Long conversations cause Claude’s performance to degrade—a phenomenon called context rot—leading to hallucinations and slower responses.

The solution is to periodically prune or summarize the conversation, keeping the context window lean for consistent quality.

Upgrade 4: Continuous Monitoring

After deployment, set up automated checks that re‑run verification and stress tests whenever code changes or new data is added.

This ensures ongoing reliability and catches regressions before they impact customers or revenue.