fix: use stderr for all logs (HF Spaces captures stderr, not stdout)

fix: redirect all log output to stdout so HF runtime log API captures it

fix: start services before init to avoid HF startup timeout

feat: add verbose logging to sync daemon and entrypoint

clean: remove all VNC/desktop files and references

fix: dataset = exact disk mirror, minimal excludes, subcommand daemon

feat: git_sync_daemon — gitfs-like auto-sync without FUSE

fix: persist HF_TOKEN/HF_DATASET_REPO via /etc/huggingrun.env

feat: reliable filesystem persistence via upload_folder API

feat: git-based filesystem persistence (dataset = disk mirror)

feat: single-port SSH-over-WebSocket + full OS persistence

feat: Ubuntu Server with ttyd web terminal + SSH + Claude Code + persistence stress test

feat: SSH access + stress test for Ubuntu desktop; always start sshd

fix: usermod -d use user not EXISTING; add scripts/curl_logs.sh for Bearer curl

fix: noVNC listen 0.0.0.0 so HF proxy can reach; APP_STARTING+URL ready => run test

fix: useradd UID 1000 already exists in base; poll once then show logs immediately on error

fix: noVNC run in foreground with --web; dbus optional; add --wait-url for poll-without-token; doc PUSH_DEBUG

deploy: Ubuntu desktop as root Dockerfile; add push-debug workflow doc; generic monitor retries and sync retry

fix: monitor_and_test.py global declaration before use

chore: add remote logs doc + monitor script, restore Dockerfile pip step

fix: sync_hf.py try block indentation (SyntaxError)

HuggingRun: 通用接口 + HF 限制文档 + APP_PORT + ubuntu-desktop 示例

HuggingRun: Run anything on Hugging Face — generic Docker + persistence