Trails

About

Trails is a reading experiment that links passages across a library of 100 non-fiction books, mostly Hacker News favourites, into short themed paths. Each trail strings together excerpts from different books around one idea, such as "Useful Lies" (self-deception as strategy), "Invisible Crack" (tiny defects that fail catastrophically) or "Pacemaker Principle" (one constraint dictating the whole system), with highlights and notes on how each passage leads to the next. It is for curious readers who want to discover unexpected connections between books, and for anyone hunting for their next read. It was made by the independent developer behind pieterma.es, who argues that "LLMs are overused to summarise and underused to help us read deeper". His explainer describes how he first hand-assembled an LLM pipeline, then handed the job to Claude Code with a minimal prompt and debugging tools, and found it did better than his hand-tuned version; Claude Code became his main interface to the project and even proposed and implemented new commands. The Show HN post, "I used Claude Code to discover connections between 100 books", earned more than 500 points in early 2026. Highlights: • Dozens of trails, each a curated sequence of excerpts on one idea • Highlighted passages with linking notes that explain each connection • Topic tags on every trail for quick scanning • A books page with the full 100-book library and covers • A detailed public explainer of the agent pipeline: ideation, research and connection • Novelty scoring that pushes the agent toward under-explored books and topics Trails is free to read, with no account or sign-up. It is a website and works in any modern browser on desktop and mobile.

Reviews