Ship v0.1.1–v0.1.5: first-live-run fixes and dashboard viewer

- v0.1.1: config persistence — FileHelper paths made absolute
  (/media/startos/volumes/main/...); relative paths resolved into the JS
  runtime's ephemeral cwd so action saves never reached /data
- v0.1.2: preJobStopContainers (Configure Grading) — docker-stop resident
  vLLM containers on the head Spark at job start, no auto-restart
- v0.1.3: preflight auth (LiteLLM master_key gates /models),
  poll-until-loaded, crash fast-fail (restarting counts as dead)
- v0.1.4: HF_HUB_OFFLINE/TRANSFORMERS_OFFLINE in airgapped serving
  (--internal network has no DNS); grader _post timeout 600→1800s for
  ~3.6 tok/s GB10 generation
- v0.1.5: dashboard viewer survives the periodic background refresh;
  download buttons for deck reports, deck JSON, and SCORECARD.md
- .gitignore: .startos/ build workspace, start-technologies/

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>

sandbox/grader_agent.py

@@ -134,7 +134,11 @@ def write_invalid_marker(err: str) -> None:
 
 
 # ---------------------------------------------------------------- LLM client
-def _post(payload: dict, timeout: int = 600) -> dict:
+def _post(payload: dict, timeout: int = 1800) -> dict:
+    # 1800s: the DGX Sparks generate a 31B at only a few tokens/sec, and
+    # structured-output (json_schema/guided) decoding is slower still — a full
+    # extraction can legitimately run past 10 minutes. Non-streaming urlopen
+    # times out on total wait, so this must cover the whole completion.
     req = urllib.request.Request(
         f"{LLM_BASE}/chat/completions",
         data=json.dumps(payload).encode(),