Xhs Profile Archive

xhs-profile-archive

Use this skill when the user wants full-profile scraping/backup for Xiaohongshu.

Inputs

• Profile URL (must contain /user/profile/<user_id>)
• Output root path (optional, default in script)

Workflow

1. Ensure browser login cookies exist for xiaohongshu.com (local Chrome cookie store).
2. Run script:
   • python3 scripts/download_xhs_profile_full.py --profile-url "<URL>" --output-root "<DIR>"
3. Validate output:
   • summary.json exists
   • notes_index.json exists
   • notes/*/content.md, notes/*/comments.json, notes/*/images/ exist
4. Return absolute archive path and counts.
5. Mandatory memory update:
   • append scrape stability notes to ../state/YYYY-MM-DD-<topic>.md
   • if retry/fallback strategy improved, update ../LEARNINGS.md
   • commit + push to main

Output contract

• summary.json: note/image/comment totals
• notes_index.json: per-note index
• notes/<idx_noteid>/content.md
• notes/<idx_noteid>/note_detail.json
• notes/<idx_noteid>/comments.json
• notes/<idx_noteid>/images/*

Safety

• Use logged-in session only; do not bypass auth.
• If comments are partially missing, report it explicitly.