audit: fix 9 defects from the 2026-09-04 hidden-bug sweep

None of these touch the signal path. The 12-month M15 book
(2024-03-31..2025-03-30, $0.65, cached Oanda bars) is byte-identical
before and after (md5 c59d3e7e685c3c880b614640adfea74b, 375 trades),
and signal/indicator logging to logs/signals.db is unchanged — verified
end to end: per-TF indicators for all 3 TFs, 14 candles each, MTF and
reasons all still persisted.

agent_panel: the >=2-take quorum guards the CHAIR's synthesizer spend,
  but ran BEFORE the macro branch — so the shipped default
  (verdict_source="macro", which never calls the synthesizer) silently
  dropped a good macro verdict whenever the other three personas were
  slow. Quorum moved below the macro path.
analyze_signal_log: _forward_return clamped to the last available bar
  when the horizon had not elapsed, reporting a near-zero "forward move"
  for every row logged within `hours` of the end of the data — on every
  run, blended into the hit-rate/mean-move aggregate. Now returns None.
analyze_signal_log: persona_scorecard iterated raw rows through an index
  that maps EVERY member row of an event, so a re-logged scan appended
  the same event R twice. One vote per event, matching
  news_outcome_buckets. Measured inflation on the live DB: 1.03x.
signal_log: dedup required a non-negative logged_at gap, but logged_at is
  stamped before the row wins the write lock (busy_timeout 5 s), so two
  processes logging one setup at a boundary can invert timestamp order
  against id order and seed a second event. Now abs(gap); same in
  backfill_seeds. Zero occurrences in the current 366 setup rows.
advisory_outcome: the catch-all turned real evaluator faults into the
  same UNRESOLVABLE used for legitimate weekend gaps, with nothing in the
  log. Still fail-safe, now logged.
llm_client: only the "cap" warn-once flag was ever reset, so in a daemon
  the second and every later network/http/parse outage logged at DEBUG
  only. Cleared on UTC-day roll and re-armed by a recovered call.
dashboard.js: renderCards() redraws the selector for every timeframe but
  was gated on the SELECTED one changing, leaving the other buttons stale
  (up to 4 h on H4) while renderHero's rail already showed the new value.
news_summary.php: atomic_write treated a short write (disk full) as
  success, publishing a truncated body the cache-first branch then served
  forever; and the daily cap was charged before the no-key check and
  before the stampede lock re-checked the cache, so a click burst could
  exhaust the 150/day budget with zero LLM calls.
migrate_signal_log_news: logged_at is ISO with 'T' while SQLite's
  datetime() uses a space, so string-comparing them put the whole cutoff
  calendar day on the wrong side and it never pruned. NOTE: this now
  deletes one day more than before — it only runs when invoked by hand.

Tests 414 -> 422 (new Module 28, one per fix). The existing
"<2 valid takes" test was retargeted at the chair — the case it actually
protects — rather than deleted, with new coverage for macro-alone.

Not fixed, recorded for the owner:
- Gate 6c's boost condition is textually identical to the C-grade
  definition, so every C in a "transitioning" regime is promoted to B
  before gate 6e reads it, making sell_min_quality's exclusion of "C" a
  no-op on M15/H4 (H1 and M5 are immune by other means). The backtester
  runs the same code, so every pinned book already includes this —
  changing it is a rule-1 revalidation event.
- Safe-haven/geopolitical headlines score exactly 0.00 in the keyword
  fallback and are now dropped entirely by the ±0.1 neutral filter,
  contradicting the platform's own gold doctrine. Bites only when
  DeepSeek is capped/disabled/failing.

Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01Ah5EB1goi5inQmSd37Ne6a

analyze_signal_log.py

@@ -444,6 +444,14 @@ def _forward_return(
     """Close-to-close move from `price` to the last close at ts+hours."""
     target = pd.Timestamp(ts) + pd.Timedelta(hours=hours)
     closes = df["Close"]
+    # The horizon must have ELAPSED. Without this the searchsorted below
+    # clamps to the last available bar, so every row logged within
+    # `hours` of the end of the data reported the latest close as its
+    # "forward" move — a near-zero reading for a trade that has had no
+    # time to develop, blended into the hit-rate/mean-move aggregate on
+    # every run. Not-yet-known is None, the same as insufficient history.
+    if len(df.index) == 0 or df.index[-1] < target:
+        return None
     pos = int(df.index.searchsorted(target, side="right")) - 1
     if pos < 0:
         return None
@@ -753,6 +761,7 @@ def persona_scorecard(rows: Sequence[LogRow], events: Sequence[SignalEvent],
     by_row = _row_index(events)
     con = open_readonly(db_path)
     agg: Dict[Tuple[str, str], List[float]] = {}
+    seen_events: set = set()
     try:
         for rid, takes in con.execute(
                 "SELECT id, panel_takes_json FROM signals"
@@ -760,6 +769,15 @@ def persona_scorecard(rows: Sequence[LogRow], events: Sequence[SignalEvent],
             ev = by_row.get(rid)
             if ev is None or ev.r_headline is None:
                 continue
+            # One vote per EVENT, not per row. A scan re-logged inside the
+            # cluster window carries its own panel takes, and counting
+            # both rows appended the SAME event R twice — inflating n and
+            # the persona means (news_outcome_buckets already iterates
+            # unique events; this is the same rule).
+            ekey = ev.rows[0].id if ev.rows else rid
+            if ekey in seen_events:
+                continue
+            seen_events.add(ekey)
             try:
                 for t in json.loads(takes):
                     agg.setdefault((t.get("persona", "?"),