fix: add SKILLOPT_SLEEP_PYTHON override + lookback_hours first-run fallback

Two fixes from issue #57 feedback:

1. run-sleep.sh: support SKILLOPT_SLEEP_PYTHON env var to explicitly set
   the Python interpreter. Useful on macOS where system Python is 3.9 but
   a newer Python is available elsewhere (e.g. Codex Desktop's bundled
   Python 3.12). Applied to both the shared runner and the bundled
   Claude Code plugin copy.

2. cycle.py: on first run (no prior harvest recorded), apply the
   lookback_hours config (default 72h) as a time cutoff. Previously,
   first run scanned the entire transcript history, which could trigger
   massive LLM mining on users with months of session data.

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

skillopt_sleep/cycle.py

@@ -144,6 +144,14 @@ def run_sleep_cycle(
         _progress(cfg, f"using {len(tasks)} seeded tasks")
     else:
         since = state.last_harvest_for(project)
+        # On first run (no prior harvest), apply lookback_hours so we don't
+        # scan the entire transcript history and trigger massive LLM mining.
+        if since is None:
+            lookback_hours = cfg.get("lookback_hours", 72)
+            if lookback_hours and lookback_hours > 0:
+                import time
+                cutoff = time.time() - lookback_hours * 3600
+                since = _now_iso(cutoff)
         max_tasks = cfg.get("max_tasks_per_night", 40)
         max_sessions = cfg.get("max_sessions_per_night", 0) or max_tasks * 3
         candidate_limit = max_tasks