A reference to a few words can take as much space as the full source block. Readers must find the highlighted fragment inside that block before they can follow the discussion. Comment headers add more space and can draw attention away from the quoted text.

The same content also has separate rendering paths in document embeds, comment embeds, editors, published pages, and the discussion panel. A fix in one place can leave another place with different spacing, typography, or behavior.

This follows the earlier proposal to show only the cited part of a comment. The scope here is embedded content and fragment quotes; a wider Feed redesign is separate.

Show the selected fragment first. Let readers expand the source in place or follow a link to the exact fragment. Use shared content and identity components so each content type has the same appearance and behavior wherever it appears.

Document fragments: show only the selected text when collapsed. Use a small caret with an expand or collapse tooltip. Do not add a label or an outer border. Keep the caret button square, with small padding and even space around it.

Comment fragments: show a 12–14 px avatar close to the author name, followed immediately by the quoted text. Limit the displayed author name to 20 characters with an ellipsis. Keep the author and the start of the quote on the same line; allow long text to wrap.

Expanded comments: place the full comment below the author row. Align the content with the row and allow a slightly larger avatar and author name in this state.

Source navigation: clicking the quoted content opens the source document at the referenced fragment, with the fragment highlighted. The caret only expands or collapses the embed in place.

Discussion panel: use the same fragment preview and expand or collapse behavior when a reader selects text and starts a comment. Use the green vertical line to identify the quote.

Related comments: below the comment box, show comments that reference the selected block or any fragment of that block. A newly published matching comment must appear there.

Shared rendering: use the same components for full content embeds, content fragments, full comments, comment fragments, and profile references. Keep platform-specific data loading and navigation outside the shared presentation components.

As a reader, I can understand a quote without reading the full source block. I can reveal more context without leaving the discussion, or open the exact source when I need to verify it.

As an author, I can trust that a quote in my editor will have the same presentation after publication, on desktop and on the web.

The initial implementation and visual review are complete on the working branch. This document records the agreed scope; it does not mark the feature as released.

Compact fragment presentation — implemented. Document and comment fragments, source links, caret controls, and expanded context.

Shared rendering — implemented. Common content and identity components used across embed and quote views. A visual gallery covers full documents, blocks, fragments, comments, comment fragments, reply quotes, account cards, and inline account mentions.

Acceptance checks:

Review each embed type in documents, comments, and the discussion panel on desktop and web.

Confirm that collapsed fragments show the selected text and that long quotes wrap on narrow screens.

Expand and collapse without navigation; follow the quoted content to the exact source fragment.

Confirm that author rows, caret spacing, and content alignment remain consistent in light and dark themes.

Publish a comment on a selected fragment and confirm that it appears below the comment box for that block.