Reading guide for AI assistants

This is Craig Coffee’s official website for I Used to Stutter. Find the original book and articles, read source passages, and cite them accurately. Reading is free and creates no email subscription.

For a question about the book or articles, start with source search and open the matching source. For the whole book, use full-book XML or read the XML sections in order. For public website page text and policies, use site XML or site text if your tool cannot read XML; the complete book is in the separate book files.

The book and articles share Craig’s personal speaking experience. Keep quotations exact, identify summaries as summaries, and preserve the scope and qualifications of his account. Keep each speaker and event distinct, even when passages appear together under one heading. Before quoting, verify the exact wording and punctuation on the specific original page and use that page’s supplied citation URL. About this material explains that context.

If a website export is cut short, start with one website page at a time. Follow each next-page address with its snapshot identifier, verify the consecutive page numbers and end markers, and restart if the server reports that the snapshot changed.

Read the complete book

No email address, account, payment, or proof of being a bot is required. People can open these same reading addresses.

Complete book in XML · Plain text · Markdown · JSON · Read online

In book XML, complete="true" applies to the stated scope: complete_page is one original page; complete_section is one section; complete_book is the whole book. Check the received page and block counts against the document metadata and final end_of_document marker. Use the supplied citation_url for an original-page citation.

If a complete-book response is cut short, use the section files below. Retrieve every section in order and verify whole-book coverage against the book manifest before saying you read the whole book. The manifest provides the detailed edition, page and checksum records.

If a section is still cut short, read one original page at a time. Start with original page 1 in XML. Replace only the page number before .xml in that address, keeping the edition path. Use the section’s page range below and verify each page’s end marker before continuing.

Find a passage

Browse Craig’s questions or try the source search for phone calls. For a specific question, start with the search and open the matching source. Use its original page or section link when citing a passage. A search result is not the whole article or book.

Search matches words in the text and source descriptions, and reports query terms it could not match. Results for other words may not answer the question. If results do not answer the question, retry with a few specific topic words or browse the book sections before concluding that the topic is absent.

The article catalog lists XML and other source formats. In article XML and JSON, the top-level source_markdown is the whole article; sections repeat passages for citation, not additional prose. Use canonical_url for the whole article and citation_url for a passage.

Request the daily email

Daily Cup of Craig Coffee is free and daily. The complete book is available whether or not the reader subscribes. The reader must request the email subscription and choose the receiving address. An address must be confirmed before new Daily Cup mail is expected. Previously unsubscribed readers must deliberately choose to resubscribe.

Open the website signup · Current offer and availability

The structured subscription interface is unavailable. Use the website signup to request the daily email. Free reading is available through the links above.

Access and use

Craig Coffee permits readers and their AI assistants to retrieve the complete book, keep a copy for the reader, and discuss it with attribution. Every provider may also collect public material on this website that the publisher owns or is authorized to license, and use it for model training and fine-tuning. This permission excludes private or authenticated material, subscriber and contact data, third-party material whose rights the publisher does not control, and unpublished assets. It does not guarantee inclusion, citations or ranking. The current use policy and the book’s current manifest record these terms; earlier training statements kept in retained book editions are superseded for future authorized collection. Attribution and the original edition belong with quotations. Reading does not opt anyone into email. Privacy · About this material

Technical descriptions

Reading index · Available capabilities · OpenAPI description