Reverse-engineer a book file format to search it
After the RAG indexer crashed on an uploaded book, Claude used Unix commands to work out the file format and grep the book manually.
Done withClaude
- What they did
- The commenter uploaded a book into the project context on Claude's web chat. The RAG indexer crashed on the file. Claude then used a series of Unix commands to work out the book's file format, and searched the contents with grep by hand. The source doesn't say whether the commenter prompted any of these steps.
- How it went
- Claude got around the crashed indexer and could search the book anyway. The commenter found this surprising. The source gives no details on accuracy, the file format, or how long it took.
Try it yourself with Claude Code
I have a file at [path to file] that my tools can't read properly, and it is a [book/document, format if known]. Use command-line tools to inspect its structure, work out the format, and extract the readable text into a plain text file. Then search it for [topic or phrase] and show me the matching passages with their locations. Don't modify the original file.
Source: HN · Aug 31, 2026
Five of these in your inbox every morning
The best things people got an AI agent to do, each with the prompt to try it.
More like this
Find connections across 100 non-fiction books
Claude Code browses a personal library with CLI tools and surfaces trails of related excerpts across books.
Research a blog post with Claude computer use
The author used Claude computer use with Claude in Chrome to research the post itself, suggesting starting with a logged-in site and a specific lookup.
- Custom agent
Search literature and write notes into Zotero
A researcher uses Claude to search Google Scholar, follow citations, query a Zotero library in natural language, and write summary notes into Zotero or Obsidian.
Find Reddit threads on reader pain points
A newsletter writer had Comet find Reddit threads on reader pain points, finding extra sources in about 15 minutes without intervention.
Build a personal research wiki from papers and articles
Cursor Agent writes and maintains interlinked markdown pages from user-supplied sources, summarizing and cross-referencing, with Obsidian used to browse.
Compile a daily digest of Reddit AI tool chatter
A Lindy agent searches Reddit for conversations about new AI tools, compiles popular topics and updates a document daily.