fix: AUTO tier order ascending + stats/CSV export completeness + UI key-view toggle & exit affordance

- scheduler: BuildChain sorts tiers ascending so tier 1 (highest priority) is tried first; previously descending inverted the chain (P10-1..P10-5)
- stats/api: handleStatsAPI limit→20000; CSV reads full audit via new Stats.AuditRecords (includes *.old rotation); key_names mapped by keyID() masked key; non-admin filter and keys-csv use masked keys
- server: ring buffer NewStats(10000)
- ui: dash-row card scroll area moved to table container (fix overflow below card); key view toggle (re-click same key returns to global) + kpi-exit affordance; rec-exit span
- chat: chain-failure record now surfaces first failed tier; AUTO comment sync

internal/gateway/chat.go

@@ -648,10 +648,11 @@ func (g *Gateway) streamChat(w http.ResponseWriter, ctx context.Context, cands [
 }
 
 // singleChatAuto runs a non-streaming AUTO request down the chain (see
-// scheduler.ChainChat): tiers descending, per-tier round-robin ordered by
-// preference, cooldown as the only hard skip, busy slots skipped without
-// penalty and a bounded busy wait. When every tier fails, the response is a
-// 503 carrying the per-tier error summary (which source/model failed why).
+// scheduler.ChainChat): tiers ascending (tier 1 = highest priority first),
+// per-tier round-robin ordered by preference, cooldown as the only hard skip,
+// busy slots skipped without penalty and a bounded busy wait. When every
+// tier fails, the response is a 503 carrying the per-tier error summary
+// (which source/model failed why).
 func (g *Gateway) singleChatAuto(w http.ResponseWriter, ctx context.Context, chain *scheduler.Chain, req *types.ChatRequest, rec *Req, quotaExhausted func(*scheduler.Slot) bool) {
 	rec.LatMs = 0
 	t0 := time.Now()