How to Write CLAUDE.md Files That Actually Work (Harness Engineering #2)

Source: Dev.to

Findings from ETH Zurich

Researchers tested 138 agent files across multiple AI coding agents. The results:

• Human‑written, concise (< 60 lines): +4 % success rate
• LLM‑generated, verbose (200 + lines): ‑3 % success rate, +20 % token cost

LLM‑generated files made agents worse.

What This Guide Covers

• The 60‑line principle: what to include, what to leave out
• Anti‑pattern gallery: documentation dump, LLM manifesto, everything file
• Progressive disclosure with Skills
• Templates for 3 project types: monorepo, API, frontend
• How to measure if your CLAUDE.md is working

Further Reading

• Full article: How to Write CLAUDE.md Files That Actually Work (Harness Engineering #2)
• If you found this useful, check out my blog for more AI engineering guides.