- Score max changed from 100 to 60, rating max from 10 to 6
- Note scoring now awards actual rating points (delta-based)
- Re-scoring only awards/deducts the difference
- Rating history shows note_score entries with problem link
- LLM prompt includes problem statement context for better evaluation
- LLM scoring dimensions: 题意理解/思路算法/代码记录/踩坑反思 (15 each)
- Minecraft-themed UI: 矿石鉴定, 探索笔记, 存入宝典, etc.
- Fallback scoring adjusted for 60-point scale
- Handle LLM markdown code fence wrapping in response
Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>
- Problems page: replace Luogu pass rate with local submission stats
(local_submit_count, local_ac_count)
- Problems page: add user AC/fail status column (user_ac, user_fail_count)
- Admin users: add total_submissions and total_ac columns
- Admin users: add detail panel with submissions/rating/redeem tabs
- Admin: new endpoint GET /api/v1/admin/users/{id}/rating-history
- Rating history: note field includes problem title via JOIN
- Me page: translate task codes to friendly labels with icons
- Me page: problem links in rating history are clickable
- Wrong book service, learning note scoring, note image controller
- Backend SQL uses batch queries for performance
Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>
- Add SubmissionFeedbackRunner: async background queue for auto LLM feedback
- Enqueue feedback generation after each submission in submitProblem()
- Register runner in main.cc with CSP_FEEDBACK_AUTO_RUN env var
- Add problem_title to GET /api/v1/submissions/{id} response
- Frontend: clickable problem link on submission detail page
- Enhance LLM prompt with richer analysis dimensions
- Add 5xx/connection error retry (max 5 attempts) in Python LLM script
Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>
