bugfixes

bin/create-user

@@ -137,13 +137,13 @@ echo -e "${GREEN}User ID: $USER_ID${NC}"
 # Build license JSON
 case "$LICENSE_TYPE" in
     enterprise)
-        LICENSE_JSON='{"licenseType":"enterprise","features":{"maxIndicators":200,"maxStrategies":100,"maxBacktestDays":1825,"realtimeData":true,"customExecutors":true,"apiAccess":true},"resourceLimits":{"maxConcurrentSessions":20,"maxMessagesPerDay":10000,"maxTokensPerMessage":32768,"rateLimitPerMinute":300},"k8sResources":{"memoryRequest":"1Gi","memoryLimit":"4Gi","cpuRequest":"500m","cpuLimit":"4000m","storage":"50Gi","tmpSizeLimit":"1Gi","enableIdleShutdown":true,"idleTimeoutMinutes":120},"preferredModel":{"provider":"anthropic","model":"claude-opus-4-6","temperature":0.7}}'
+        LICENSE_JSON='{"licenseType":"enterprise","features":{"maxIndicators":999,"maxStrategies":999,"maxBacktestDays":3650,"realtimeData":true,"customExecutors":true,"apiAccess":true},"resourceLimits":{"maxConcurrentSessions":20,"maxMessagesPerDay":10000,"maxTokensPerMessage":32768,"rateLimitPerMinute":300},"k8sResources":{"memoryRequest":"1Gi","memoryLimit":"8Gi","cpuRequest":"500m","cpuLimit":"4000m","storage":"50Gi","tmpSizeLimit":"512Mi","enableIdleShutdown":false,"idleTimeoutMinutes":0}}'
         ;;
     free)
-        LICENSE_JSON='{"licenseType":"free","features":{"maxIndicators":10,"maxStrategies":3,"maxBacktestDays":30,"realtimeData":false,"customExecutors":false,"apiAccess":false},"resourceLimits":{"maxConcurrentSessions":1,"maxMessagesPerDay":100,"maxTokensPerMessage":4096,"rateLimitPerMinute":20},"k8sResources":{"memoryRequest":"256Mi","memoryLimit":"512Mi","cpuRequest":"100m","cpuLimit":"500m","storage":"2Gi","tmpSizeLimit":"128Mi","enableIdleShutdown":true,"idleTimeoutMinutes":30},"preferredModel":{"provider":"anthropic","model":"claude-haiku-4-5-20251001","temperature":0.7}}'
+        LICENSE_JSON='{"licenseType":"free","features":{"maxIndicators":5,"maxStrategies":3,"maxBacktestDays":30,"realtimeData":false,"customExecutors":false,"apiAccess":false},"resourceLimits":{"maxConcurrentSessions":1,"maxMessagesPerDay":100,"maxTokensPerMessage":4096,"rateLimitPerMinute":10},"k8sResources":{"memoryRequest":"256Mi","memoryLimit":"8Gi","cpuRequest":"100m","cpuLimit":"500m","storage":"1Gi","tmpSizeLimit":"128Mi","enableIdleShutdown":true,"idleTimeoutMinutes":15}}'
         ;;
     pro|*)
-        LICENSE_JSON='{"licenseType":"pro","features":{"maxIndicators":50,"maxStrategies":20,"maxBacktestDays":365,"realtimeData":true,"customExecutors":true,"apiAccess":true},"resourceLimits":{"maxConcurrentSessions":5,"maxMessagesPerDay":1000,"maxTokensPerMessage":8192,"rateLimitPerMinute":60},"k8sResources":{"memoryRequest":"512Mi","memoryLimit":"2Gi","cpuRequest":"250m","cpuLimit":"2000m","storage":"10Gi","tmpSizeLimit":"256Mi","enableIdleShutdown":true,"idleTimeoutMinutes":60},"preferredModel":{"provider":"anthropic","model":"claude-sonnet-4-6","temperature":0.7}}'
+        LICENSE_JSON='{"licenseType":"pro","features":{"maxIndicators":50,"maxStrategies":20,"maxBacktestDays":365,"realtimeData":true,"customExecutors":true,"apiAccess":true},"resourceLimits":{"maxConcurrentSessions":5,"maxMessagesPerDay":1000,"maxTokensPerMessage":8192,"rateLimitPerMinute":60},"k8sResources":{"memoryRequest":"512Mi","memoryLimit":"8Gi","cpuRequest":"250m","cpuLimit":"2000m","storage":"10Gi","tmpSizeLimit":"256Mi","enableIdleShutdown":false,"idleTimeoutMinutes":0}}'
         ;;
 esac
 

deploy/k8s/dev/configs/flink-config.yaml

@@ -2,7 +2,7 @@
 
 # ZeroMQ bind address and ports
 zmq_bind_address: "tcp://*"
-zmq_ingestor_work_queue_port: 5555
+zmq_ingestor_work_queue_port: 5567
 zmq_market_data_pub_port: 5558
 
 # Notification endpoints

deploy/k8s/prod/configs/flink-config.yaml

@@ -2,7 +2,7 @@
 
 # ZeroMQ bind address and ports
 zmq_bind_address: "tcp://*"
-zmq_ingestor_work_queue_port: 5555
+zmq_ingestor_work_queue_port: 5567
 zmq_market_data_pub_port: 5558
 
 # Notification endpoints (internal Flink task manager → job manager path)

deploy/k8s/prod/configs/gateway-config.yaml

@@ -21,6 +21,30 @@ data:
       model_provider: deepinfra
       model: zai-org/GLM-5
 
+    # License tier model configuration
+    license_models:
+      # Free tier models
+      free:
+        default: zai-org/GLM-5
+        cost_optimized: zai-org/GLM-5
+        complex: zai-org/GLM-5
+        allowed_models:
+          - zai-org/GLM-5
+
+      # Pro tier models
+      pro:
+        default: zai-org/GLM-5
+        cost_optimized: zai-org/GLM-5
+        complex: zai-org/GLM-5
+        blocked_models:
+          - Qwen/Qwen3-235B-A22B-Instruct-2507
+
+      # Enterprise tier models
+      enterprise:
+        default: zai-org/GLM-5
+        cost_optimized: zai-org/GLM-5
+        complex: Qwen/Qwen3-235B-A22B-Instruct-2507
+
     # Kubernetes configuration
     kubernetes:
       namespace: sandbox
@@ -40,10 +64,17 @@ data:
       url: http://qdrant:6333
       collection: gateway_memory
 
+    # Agent configuration
+    agent:
+      # Number of prior conversation turns loaded as LLM context and flushed to Iceberg at session end
+      conversation_history_limit: 20
+
     # Iceberg (for durable storage via REST catalog)
     iceberg:
       catalog_uri: http://iceberg-catalog:8181
       namespace: gateway
+      ohlc_catalog_uri: http://iceberg-catalog:8181
+      ohlc_namespace: trading
       s3_endpoint: http://minio:9000
       conversations_bucket: warehouse
 

deploy/k8s/prod/configs/ingestor-config.yaml

@@ -10,12 +10,25 @@ supported_exchanges:
   - COINBASE
   - KRAKEN
 
+# Per-exchange work slot capacity.
+# Each slot is one concurrent job. historical_slots limits parallel OHLC fetches;
+# realtime_slots limits concurrent tick subscriptions.
+exchange_capacity:
+  BINANCE:
+    historical_slots: 1
+    realtime_slots: 5
+  COINBASE:
+    historical_slots: 1
+    realtime_slots: 4
+  KRAKEN:
+    historical_slots: 1
+    realtime_slots: 3
+
 # Kafka configuration
 kafka_brokers:
   - kafka:9092
 
 # Worker configuration
-max_concurrent: 10
 poll_interval_ms: 10000
 
 # Logging

deploy/k8s/prod/kustomization.yaml

@@ -15,9 +15,8 @@ resources:
   - infrastructure.yaml
   # Sandbox namespace resources (go to sandbox namespace, not ai)
   - sandbox-config.yaml
-  # gateway-config ConfigMap is intentionally excluded from kustomize.
+  # gateway-config ConfigMap (database URL is in secrets, not here)
-  # It contains an op:// reference for the DB password. Apply via:
+  - configs/gateway-config.yaml
-  #   bin/config-update prod gateway-config
 
 patches:
   - path: patch-gateway-rbac-subject.yaml

doc/plan.md

@@ -1,5 +1,6 @@
 # Development Plan
 
+* Single conversation in gateway
 * Realtime data
 * Triggers
 * Strategy UI
@@ -8,3 +9,8 @@
 * User secrets
 * Live Execution
 * Sandbox <=> Dexorder auth
+* Chat channels
+* MCP channel (with or without images)
+* TradingView indicator import tool
+* Trader preferences tool
+* 

doc/prod_deployment.md

@@ -30,6 +30,16 @@ This script (hardcoded to `--context=prod`) performs:
 
 > **Secrets are NOT updated by this script.** Run `bin/secret-update prod` separately if secrets have changed.
 
+### Post-deploy: refresh user licenses
+
+After any deploy that changes license tier templates (`gateway/src/types/user.ts`), run:
+
+```bash
+bin/create-all-users prod
+```
+
+This upserts all alpha users and re-applies the current tier template to their `user_licenses` row. Safe to run on an existing database — it will not delete users or lose data. New sandbox deployments will pick up the updated resource limits on next login.
+
 ---
 
 ## Full Deploy with Iceberg Schema Wipe
@@ -137,3 +147,20 @@ kubectl --context prod -n ai logs deployment/gateway --tail=100
 ### Gateway shows `42P01` errors but pod is running
 
 The gateway does not auto-migrate on startup. The schema file must be applied manually after any database recreation. A gateway restart alone will not fix this.
+
+### Gateway CrashLoopBackOff — `ECONNREFUSED postgresql://localhost/dexorder`
+
+**Symptom:** New gateway pod crashes immediately with `Database connection failed` and logs show `databaseUrl: "postgresql://localhost/dexorder"`.
+
+**Cause:** The gateway reads `database.url` from `config.yaml` (via `configData`). If that key is absent, it falls back to the localhost default — even if `secrets.yaml` has `database.url`. The code checks `configData.database?.url || secretsData.database?.url || ...` (as of `c8fa99c`), so both sources work, but both files must be present and correctly mounted.
+
+**What to check:**
+1. Does the `gateway-config` ConfigMap have a `database:` section? (It should not — credentials belong in secrets as of the nautilus branch.)
+2. Does `gateway-secrets` have `database.url`? Verify: `kubectl --context prod -n ai get secret gateway-secrets -o jsonpath='{.data.secrets\.yaml}' | base64 -d`
+3. If the secret is missing the database section, run `bin/secret-update prod` (requires 1Password desktop to be unlocked — must run interactively, not via pipe).
+
+### `bin/secret-update prod` fails with "authorization prompt dismissed"
+
+1Password's `op inject` requires interactive desktop authentication. Running it via `echo "yes" | bin/secret-update prod` or any background/piped invocation will fail silently (the script prints `✓` even though `kubectl apply` received empty input).
+
+**Fix:** Run `bin/secret-update prod` in an interactive terminal with 1Password unlocked.

gateway/src/auth/authenticator.ts

@@ -129,7 +129,7 @@ export class Authenticator {
         'Container is ready'
       );
 
-      const sessionId = `tg_${telegramUserId}_${Date.now()}`;
+      const sessionId = `tg_${telegramUserId}`;
 
       return {
         userId,

gateway/src/channels/telegram-handler.ts

@@ -248,7 +248,7 @@ export class TelegramHandler {
    * Clean up sessions that have been idle longer than maxAgeMs.
    * Triggers Iceberg flush for each expired session via harness.cleanup().
    */
-  async cleanupSessions(maxAgeMs = 30 * 60 * 1000): Promise<void> {
+  async cleanupSessions(maxAgeMs = 2 * 60 * 60 * 1000): Promise<void> {
     const now = Date.now();
     const expired: string[] = [];
 

gateway/src/channels/websocket-handler.ts

@@ -511,13 +511,18 @@ export class WebSocketHandler {
       hasSymbolIndexService: !!symbolIndexService
     }, 'Service availability');
 
+    const requestId = payload.request_id || randomUUID();
+
     if (!ohlcService && !symbolIndexService) {
-      logger.warn('No datafeed services available');
+      logger.warn({ requestId }, 'No datafeed services available yet');
+      socket.send(JSON.stringify({
+        type: 'error',
+        request_id: requestId,
+        error_message: 'Services initializing, please retry',
+      }));
       return;
     }
 
-    const requestId = payload.request_id || randomUUID();
-
     try {
       switch (payload.type) {
         case 'get_config': {

gateway/src/harness/memory/checkpoint-saver.ts

@@ -14,7 +14,7 @@ import type { FastifyBaseLogger } from 'fastify';
  * Iceberg for durable storage with time-travel capabilities.
  */
 export class TieredCheckpointSaver extends BaseCheckpointSaver<number> {
-  private readonly HOT_TTL_SECONDS = 3600; // 1 hour
+  private readonly HOT_TTL_SECONDS = 86400; // 24 hours
   private readonly KEY_PREFIX = 'ckpt:';
 
   constructor(

gateway/src/harness/memory/conversation-store.ts

@@ -30,7 +30,7 @@ export interface StoredMessage {
  */
 export class ConversationStore {
   private readonly HOT_MESSAGE_LIMIT = 50; // Redis buffer ceiling
-  private readonly HOT_TTL_SECONDS = 3600; // 1 hour
+  private readonly HOT_TTL_SECONDS = 86400; // 24 hours
 
   constructor(
     private redis: Redis,

ingestor/src/index.js

@@ -372,7 +372,10 @@ class IngestorWorker {
 
       if (candles.length > 0) {
         const metadata = { request_id: requestId, client_id, ticker, period_seconds, start_time, end_time };
-        const PAGE_SIZE = 1000;
+        // 8000 rows/page: each OHLC row is ~77 bytes typical (9 populated fields as
+        // protobuf varints + ticker string). Worst-case is ~124 bytes, so 8000 rows
+        // stays safely under Kafka's 1MB message limit in all realistic scenarios.
+        const PAGE_SIZE = 8000;
         for (let i = 0; i < candles.length; i += PAGE_SIZE) {
           const page = candles.slice(i, i + PAGE_SIZE);
           const isLastPage = (i + PAGE_SIZE) >= candles.length;

ingestor/src/symbol-metadata-generator.js

@@ -249,7 +249,6 @@ export class SymbolMetadataGenerator {
 
       if (!this.publishedSymbols.has(key)) {
         uniqueMetadata.push(metadata);
-        this.publishedSymbols.add(key);
       } else {
         duplicateCount++;
       }
@@ -275,6 +274,12 @@ export class SymbolMetadataGenerator {
 
     await this.kafkaProducer.writeMarketMetadata(topic, messages);
 
+    // Mark as published only after successful Kafka write
+    for (const metadata of uniqueMetadata) {
+      const key = `${metadata.marketId}.${metadata.exchangeId}`;
+      this.publishedSymbols.add(key);
+    }
+
     this.logger.info(
       { count: messages.length, duplicateCount, topic },
       'Wrote symbol metadata to Kafka'

web/src/composables/useTradingViewDatafeed.ts

@@ -45,6 +45,9 @@ export class WebSocketDatafeed implements IBasicDataFeed {
   private configuration: DatafeedConfiguration | null = null
   private messageHandler: MessageHandler
   private symbolDenominators: Map<string, SymbolDenominators> = new Map() // Track denominators per symbol
+  // Tracks the last bar time (ms) returned by getBars per "symbolKey_periodSeconds".
+  // bar_updates with time < this watermark are stale and already covered by history.
+  private lastBarTimes: Map<string, number> = new Map()
 
   constructor() {
     // Use the shared WebSocket connection (managed by App.vue authentication)
@@ -118,6 +121,14 @@ export class WebSocketDatafeed implements IBasicDataFeed {
         const symbolKey = subscription.symbolInfo.ticker || subscription.symbolInfo.name
         const denoms = this.symbolDenominators.get(symbolKey) || { tick: 1, base: 1 }
 
+        // Drop bars already covered by getBars history to prevent time-order violations
+        const barTimeMs = message.bar.time * 1000
+        const barKey = `${symbolKey}_${message.period_seconds}`
+        const watermark = this.lastBarTimes.get(barKey) ?? 0
+        if (barTimeMs < watermark) {
+          return
+        }
+
         const bar: Bar = {
           time: message.bar.time * 1000, // Convert to milliseconds
           open: parseFloat(message.bar.open) / denoms.tick,
@@ -284,6 +295,16 @@ export class WebSocketDatafeed implements IBasicDataFeed {
 
           bars.sort((a, b) => a.time - b.time)
 
+          // Update last-bar watermark so bar_update handler can drop stale replays
+          if (bars.length > 0) {
+            const barKey = `${symbolKey}_${intervalToSeconds(resolution)}`
+            const newLast = bars[bars.length - 1].time
+            const prevLast = this.lastBarTimes.get(barKey) ?? 0
+            if (newLast > prevLast) {
+              this.lastBarTimes.set(barKey, newLast)
+            }
+          }
+
           console.log('[TradingView Datafeed] Scaled bar sample:', bars[0])
 
           const meta: HistoryMetadata = {