xiaohongshu link is incorrect

[runny4184]

Routes

/xiaohongshu/user/:user_id/:category/:routeParams?

Full routes

/xiaohongshu/user/:user_id/:category/:routeParams?

Related documentation

https://docs.rsshub.app/routes/xiaohongshu

What is expected?

use /explore/ url for each article and also insert a url/iframe in the content if it's a video

What is actually happening?

xiaohongshu rss is always using profile url /profile/ instead of per article url /explore/

Deployment information

Self-hosted

Deployment information (for self-hosted)

No response

Additional info

no

This is not a duplicated issue

• I have searched existing issues to ensure this bug has not already been reported

[github-actions]

This issue might be a duplicate of existing issues. Please check:

• 小红书-用户笔记和专辑 路由 item link 有误 #5217: 相同问题 - item 链接均为 /user/profile/undefined 而非每条笔记的 /explore/xxx 链接

Feel free to ignore if none of these address your specific case.

[github-actions]

Searching for maintainers:

• /xiaohongshu/user/:user_id/:category/:routeParams?: @lotosbin @howerhe @rien7 @dddaniel1 @pseudoyu

To maintainers: if you are not willing to be disturbed, list your username in scripts/workflow/test-issue/call-maintainer.js. In this way, your username will be wrapped in an inline code block when tagged so you will not be notified.

If all routes can not be found, the issue will be closed automatically. Please use NOROUTE for a route-irrelevant issue or leave a comment if it is a mistake.
如果所有路由都无法匹配，issue 将会被自动关闭。如果 issue 和路由无关，请使用 NOROUTE 关键词，或者留下评论。我们会重新审核。

[dosubot]

This behavior depends on whether you have the XIAOHONGSHU_COOKIE environment variable configured in your deployment:

• With XIAOHONGSHU_COOKIE set: Item links use the correct https://www.xiaohongshu.com/explore/{id} format [1].
• Without the cookie (fallback path): The route falls back to a Puppeteer-based scraping path, which generates links using the https://www.xiaohongshu.com/user/profile/{userId}/{noteId} format instead [2].

If you're seeing /profile/ URLs, you're likely hitting the non-cookie fallback. Setting the XIAOHONGSHU_COOKIE environment variable should give you the /explore/ links you expect.

Regarding video content — when the cookie path is used, the route already embeds videos as HTML5 <video> elements with multiple source formats (av1, h264, h265, h266) and a poster image [3]. This should work in most RSS readers that support HTML content. The fallback path may not include as rich video embedding.

To reply, just mention @dosu.

---

Docs are dead. Just use Dosu.

[runny4184]

this problem only appens on fallback. the cookie way doesn't really work for me.
@dosu can you write me a patch on the fallback to add video tag and fix the url?

[dosubot]

I can't make PRs or code changes directly, but I can give you the exact patch for both issues. Here's what needs to change in lib/routes/xiaohongshu/user.ts:

1. Fix the URL (line 104) — Replace the relative URL resolution with an explicit /explore/ link:

- link: new URL(noteCard.noteId || id, url).toString(),
+ link: `https://www.xiaohongshu.com/explore/${noteCard.noteId || id}`,

This also applies to the renderCollect function (line 123):

- link: `${url}/${item.note_id}`,
+ link: `https://www.xiaohongshu.com/explore/${item.note_id}`,

[1]

2. Add video support — This is trickier. The fallback path's scraped data from the user profile page doesn't include video metadata or note type [2]. To get video content, each note needs to be individually fetched via getFullNote(), which is exactly what the cookie path does via renderNotesFulltext() [3].

You could adapt the fallback renderNote to call getFullNote for each note, similar to how the cookie path works. Here's a rough patch for the getUserFeeds function in user.ts:

- const renderNote = (notes) =>
-     notes.flatMap((n) =>
-         n.map(({ id, noteCard }) => ({
-             title: noteCard.displayTitle,
-             link: new URL(noteCard.noteId || id, url).toString(),
-             guid: noteCard.displayTitle,
-             description: `<img src="${noteCard.cover.infoList.pop().url}" width="${noteCard.cover.width}" height="${noteCard.cover.height}"><br>${noteCard.displayTitle}`,
-             author: noteCard.user.nickname,
-             upvotes: noteCard.interactInfo.likedCount,
-         }))
-     );
+ const urlNotePrefix = 'https://www.xiaohongshu.com/explore';
+ const renderNote = async (notes) => {
+     const promises = notes.flatMap((n) =>
+         n.map(async ({ id, noteCard }) => {
+             const link = `${urlNotePrefix}/${id}`;
+             try {
+                 const { title, description, pubDate, updated } = await getFullNote(link, false);
+                 return {
+                     title,
+                     link,
+                     description,
+                     author: noteCard.user.nickname,
+                     guid: `${urlNotePrefix}/${noteCard.noteId}`,
+                     pubDate,
+                     updated,
+                     upvotes: noteCard.interactInfo.likedCount,
+                 };
+             } catch {
+                 // Fallback to basic rendering if full fetch fails
+                 return {
+                     title: noteCard.displayTitle,
+                     link,
+                     guid: noteCard.displayTitle,
+                     description: `<img src="${noteCard.cover.infoList.pop().url}" width="${noteCard.cover.width}" height="${noteCard.cover.height}"><br>${noteCard.displayTitle}`,
+                     author: noteCard.user.nickname,
+                     upvotes: noteCard.interactInfo.likedCount,
+                 };
+             }
+         })
+     );
+     return Promise.all(promises);
+ };

And update the return statement (line 135) since renderNote is now async:

- item: category === 'notes' ? renderNote(notes) : renderCollect(collect),
+ item: category === 'notes' ? await renderNote(notes) : renderCollect(collect),

You'll also need to add getFullNote to the import on line 10:

- import { getUser, getUserWithCookie, renderNotesFulltext } from './util';
+ import { getFullNote, getUser, getUserWithCookie, renderNotesFulltext } from './util';

Important caveat: Calling getFullNote for every note without a cookie means each note page will be fetched individually via proxy or directly — this could be slow and may trigger anti-crawler protections. The cookie path works better because the cookie authenticates those requests. Without it, you might hit rate limits. A middle-ground approach would be to only fix the URL (patch #1) and skip the video embedding if cookie-less operation is a hard requirement.

To reply, just mention @dosu.

---

Docs are dead. Just use Dosu.