Oct 18 26 days ago

1. 11:11

   pod-onstart: make 160 GiB host-RAM gate opt-in (FN_MIN_RAM_GB, default 0 = no cap)⁠

   rayray916
   Verified signature

   a7ebeb1
2. 09:49

   Serve bare HTTP over tailnet; keep --ephemeral removal

   rayray916
   162b3da
3. 09:30

   Add RAM gate (>=160GiB) and Vast-safe NIXL toml patch

   rayray916
   b76a52d
4. 08:47

   Pin tailscale node name to flash-next via TS_HOSTNAME

   rayray916
   7770054
5. 08:43

   Patch image ts-up.sh: drop removed --ephemeral flag so tailscale can join

   rayray916
   b199f88
6. 08:40

   Add probe.sh: power + real HF throughput pre-rent check

   rayray916
   60f3bbc
7. 08:25

   Keep env HF_TOKEN for authenticated pulls; retry Sharp fetch anonymously on failure

   rayray916
   9cc9868
8. 08:20

   Drop inherited HF_TOKEN for anonymous pulls

   rayray916
   59ae139
9. 08:20

   Add Flash-Next aio pod bootstrap (nvidia-smi + 600W gate + Sharp template)

   rayray916
   7280dcb
10. 08:20

    initial commit

    rayray916
    Verified signature

    8a7cc8e