feat: infinite scroll on shared folder/project pages

[vishwajeet-13]

What

The guest share page (/vms/shared/folder/…, /vms/shared/…) used Previous / Next buttons — clicking through a page at a time. This swaps that for infinite scroll.

How

Reuses the exact pattern the project/folder grid already runs (useProjectBrowser + useInfiniteScroll):

• useInfiniteScroll(sentinel, loadingMore, hasMore, loadMore) — an IntersectionObserver on a sentinel <div> at the end of the grid (rootMargin: 600px)
• a request ref tracking { page, append }; loadMore() bumps the page and calls assetsCall.reload()
• onSuccess accumulates into assets with name-dedupe; reachedEnd when a page adds nothing
• the pager markup is replaced by a small spinner that only renders while hasMore

No backend change — get_shared_folder_assets / get_shared_project_assets already paginate.

Verified (local, 42-asset shared folder)

• First load: 20 cards. Scrolling fetches page 2, then page 3 → 42 cards, then stops (3 requests total, no duplicates, no console errors).
• "Download all" and the RAW/HEIC format menu (from feat: JPEG/PNG download choice on shared folder/project pages #159) still work on the accumulated list.

🤖 Generated with Claude Code

[greptile-apps]

Greptile Summary

This PR replaces Previous/Next pagination on guest shared project and folder pages with sentinel-driven infinite scrolling.

• Accumulates paginated assets while deduplicating by asset name.
• Stops loading when all reported assets are present or a page contributes no new assets.
• Preserves preview and download actions over the accumulated list.
• Needs bounded error handling to prevent automatic load-more retries from looping during request failures.

Confidence Score: 4/5

This PR is not safe to merge until failed infinite-scroll requests stop automatically retrying without a bound or backoff.

Clearing the infinite-scroll busy state after every failed reload retriggers loading while the sentinel remains visible and hasMore remains true, creating a request loop during persistent failures.

Files Needing Attention: frontend/src/pages/SharedProjectPage.vue

Important Files Changed

Filename	Overview
frontend/src/pages/SharedProjectPage.vue	Implements infinite-scroll pagination correctly on successful requests, but failed load-more requests can trigger an unbounded automatic retry cycle.

Prompt To Fix All With AI

### Issue 1
frontend/src/pages/SharedProjectPage.vue:330-331
**Failed requests retry forever**

When a load-more request fails while the sentinel remains visible, this `finally` block clears `loadingMore`. The infinite-scroll watcher immediately tries again because the sentinel still intersects and `hasMore` remains true. A persistent network or server error therefore creates an unbounded retry loop that can flood the shared-assets endpoint. Handle the failure or prevent automatic retries until a new user or observer event occurs.

---

For each issue above, determine whether it is valid and should be fixed. If so, fix it directly.

_{Reviews (1): Last reviewed commit: "feat: infinite scroll on shared folder/p..." | Re-trigger Greptile}

[greptile-apps]

Failed requests retry forever

When a load-more request fails while the sentinel remains visible, this finally block clears loadingMore. The infinite-scroll watcher immediately tries again because the sentinel still intersects and hasMore remains true. A persistent network or server error therefore creates an unbounded retry loop that can flood the shared-assets endpoint. Handle the failure or prevent automatic retries until a new user or observer event occurs.

Prompt To Fix With AI

This is a comment left during a code review.
Path: frontend/src/pages/SharedProjectPage.vue
Line: 330-331

Comment:
**Failed requests retry forever**

When a load-more request fails while the sentinel remains visible, this `finally` block clears `loadingMore`. The infinite-scroll watcher immediately tries again because the sentinel still intersects and `hasMore` remains true. A persistent network or server error therefore creates an unbounded retry loop that can flood the shared-assets endpoint. Handle the failure or prevent automatic retries until a new user or observer event occurs.

---

For each issue above, determine whether it is valid and should be fixed. If so, fix it directly.