dsh-doc-share

Pasting an AI conversation out of the browser usually gives you a wall of unstructured messages—code blocks, tool calls, and images jumbled together with no title, no summary, no chapters. If you need earlier turns, you are stuck scrolling and waiting for lazy-loaded content. dsh-doc-share solves exactly this: it turns a raw conversation into a properly typeset report with a customizable cover page, content summary statistics, and chapter-by-chapter Q&A layout, so the output is something you can hand to a colleague, print, or archive rather than a pile of raw chat screenshots.

The workflow centers on a turn-selection dialog that lists every completed turn as a card (turn number, timestamp, question/answer excerpt, image and tool badges). You can search by keyword, select all, clear all, and load earlier history directly from the dialog with no page scrolling. In the generated report, code blocks are rendered as standalone blocks, tables keep their formatting, and tool calls are collapsed into summary lines. A single generation lets you download as a PNG image, a self-contained single HTML file, an A4 auto-paginated PDF, or structured Markdown. Layout width and font size are adjustable in three levels, dark theme follows system preference, and the UI automatically adapts to Chinese or English.

It suits anyone who treats AI conversations as work artifacts: researchers archiving reasoning chains into a knowledge base, developers sharing debugging sessions with teammates who have no reason to install the tool, students compiling lecture Q&A into study documents, or product teams turning design discussions into printable briefs. The conversation is over, but you still need a proper document at the end of the day—dsh-doc-share is built for exactly that.