- eveai_chat_client updated to retrieve static files from the correct (bunny.net) location when a STATIC_URL is defined.

- Defined locations for crewai crew memory. This failed in k8s.
- Redis connection for pub/sub in ExecutionProgressTracker adapted to conform to TLS-enabled connections

common/utils/execution_progress.py

@@ -4,6 +4,7 @@ from typing import Generator
 from redis import Redis, RedisError
 import json
 from flask import current_app
+import time
 
 
 class ExecutionProgressTracker:
@@ -11,22 +12,33 @@ class ExecutionProgressTracker:
 
     def __init__(self):
         try:
-            redis_url = current_app.config['SPECIALIST_EXEC_PUBSUB']
-
-            self.redis = Redis.from_url(redis_url, socket_timeout=5)
-            # Test the connection
-            self.redis.ping()
-
+            # Use shared pubsub pool (lazy connect; no eager ping)
+            from common.utils.redis_pubsub_pool import get_pubsub_client
+            self.redis = get_pubsub_client(current_app)
             self.expiry = 3600  # 1 hour expiry
-        except RedisError as e:
-            current_app.logger.error(f"Failed to connect to Redis: {str(e)}")
-            raise
         except Exception as e:
-            current_app.logger.error(f"Unexpected error during Redis initialization: {str(e)}")
+            current_app.logger.error(f"Error initializing ExecutionProgressTracker: {str(e)}")
             raise
 
     def _get_key(self, execution_id: str) -> str:
-        return f"specialist_execution:{execution_id}"
+        prefix = current_app.config.get('REDIS_PREFIXES', {}).get('pubsub_execution', 'pubsub:execution:')
+        return f"{prefix}{execution_id}"
+
+    def _retry(self, op, attempts: int = 3, base_delay: float = 0.1):
+        """Retry wrapper for Redis operations with exponential backoff."""
+        last_exc = None
+        for i in range(attempts):
+            try:
+                return op()
+            except RedisError as e:
+                last_exc = e
+                if i == attempts - 1:
+                    break
+                delay = base_delay * (3 ** i)  # 0.1, 0.3, 0.9
+                current_app.logger.warning(f"Redis operation failed (attempt {i+1}/{attempts}): {e}. Retrying in {delay}s")
+                time.sleep(delay)
+        # Exhausted retries
+        raise last_exc
 
     def send_update(self, ctask_id: str, processing_type: str, data: dict):
         """Send an update about execution progress"""
@@ -35,12 +47,6 @@ class ExecutionProgressTracker:
                                      f"{data}")
             key = self._get_key(ctask_id)
 
-            # First verify Redis is still connected
-            try:
-                self.redis.ping()
-            except RedisError:
-                current_app.logger.error("Lost Redis connection. Attempting to reconnect...")
-                self.__init__()  # Reinitialize connection
 
             update = {
                 'processing_type': processing_type,
@@ -50,7 +56,7 @@ class ExecutionProgressTracker:
 
             # Log initial state
             try:
-                orig_len = self.redis.llen(key)
+                orig_len = self._retry(lambda: self.redis.llen(key))
 
                 # Try to serialize the update and check the result
                 try:
@@ -60,13 +66,16 @@ class ExecutionProgressTracker:
                     raise
 
                 # Store update in list with pipeline for atomicity
-                with self.redis.pipeline() as pipe:
-                    pipe.rpush(key, serialized_update)
-                    pipe.publish(key, serialized_update)
-                    pipe.expire(key, self.expiry)
-                    results = pipe.execute()
+                def _pipeline_op():
+                    with self.redis.pipeline() as pipe:
+                        pipe.rpush(key, serialized_update)
+                        pipe.publish(key, serialized_update)
+                        pipe.expire(key, self.expiry)
+                        return pipe.execute()
 
-                new_len = self.redis.llen(key)
+                results = self._retry(_pipeline_op)
+
+                new_len = self._retry(lambda: self.redis.llen(key))
 
                 if new_len <= orig_len:
                     current_app.logger.error(
@@ -83,13 +92,14 @@ class ExecutionProgressTracker:
     def get_updates(self, ctask_id: str) -> Generator[str, None, None]:
         key = self._get_key(ctask_id)
         pubsub = self.redis.pubsub()
-        pubsub.subscribe(key)
+        # Subscribe with retry
+        self._retry(lambda: pubsub.subscribe(key))
 
         try:
             # First yield any existing updates
-            length = self.redis.llen(key)
+            length = self._retry(lambda: self.redis.llen(key))
             if length > 0:
-                updates = self.redis.lrange(key, 0, -1)
+                updates = self._retry(lambda: self.redis.lrange(key, 0, -1))
                 for update in updates:
                     update_data = json.loads(update.decode('utf-8'))
                     # Use processing_type for the event
@@ -98,7 +108,13 @@ class ExecutionProgressTracker:
 
             # Then listen for new updates
             while True:
-                message = pubsub.get_message(timeout=30)  # message['type'] is Redis pub/sub type
+                try:
+                    message = pubsub.get_message(timeout=30)  # message['type'] is Redis pub/sub type
+                except RedisError as e:
+                    current_app.logger.warning(f"Redis pubsub get_message error: {e}. Continuing...")
+                    time.sleep(0.3)
+                    continue
+
                 if message is None:
                     yield ": keepalive\n\n"
                     continue
@@ -111,4 +127,7 @@ class ExecutionProgressTracker:
                     if update_data['processing_type'] in ['Task Complete', 'Task Error']:
                         break
         finally:
-            pubsub.unsubscribe()
+            try:
+                pubsub.unsubscribe()
+            except Exception:
+                pass

common/utils/redis_pubsub_pool.py

@@ -0,0 +1,84 @@
+import ssl
+from typing import Dict, Any
+
+import redis
+from flask import Flask
+
+
+def _build_pubsub_redis_config(app: Flask) -> Dict[str, Any]:
+    """Build Redis ConnectionPool config for the pubsub/EPT workload using app.config.
+    Does not modify cache or session pools.
+    """
+    cfg = app.config
+
+    config: Dict[str, Any] = {
+        'host': cfg['REDIS_URL'],
+        'port': cfg['REDIS_PORT'],
+        'db': int(cfg.get('REDIS_SPECIALIST_EXEC_DB', '0')),
+        'max_connections': int(cfg.get('REDIS_PUBSUB_MAX_CONNECTIONS', 200)),
+        'retry_on_timeout': True,
+        'socket_keepalive': True,
+        'socket_keepalive_options': {},
+        'socket_timeout': float(cfg.get('REDIS_PUBSUB_SOCKET_TIMEOUT', 10.0)),
+        'socket_connect_timeout': float(cfg.get('REDIS_PUBSUB_CONNECT_TIMEOUT', 3.0)),
+    }
+
+    # Authentication if present
+    un = cfg.get('REDIS_USER')
+    pw = cfg.get('REDIS_PASS')
+    if un and pw:
+        config.update({'username': un, 'password': pw})
+
+    # TLS when configured
+    cert_path = cfg.get('REDIS_CA_CERT_PATH')
+    if cfg.get('REDIS_SCHEME') == 'rediss' and cert_path:
+        config.update({
+            'connection_class': redis.SSLConnection,
+            'ssl_cert_reqs': ssl.CERT_REQUIRED,
+            'ssl_check_hostname': cfg.get('REDIS_SSL_CHECK_HOSTNAME', True),
+            'ssl_ca_certs': cert_path,
+        })
+
+    return config
+
+
+def create_pubsub_pool(app: Flask) -> redis.ConnectionPool:
+    """Create and store the dedicated pubsub ConnectionPool in app.extensions."""
+    if not hasattr(app, 'extensions'):
+        app.extensions = {}
+
+    # Reuse existing if already created
+    pool = app.extensions.get('redis_pubsub_pool')
+    if pool is not None:
+        return pool
+
+    config = _build_pubsub_redis_config(app)
+    pool = redis.ConnectionPool(**config)
+    app.extensions['redis_pubsub_pool'] = pool
+
+    # Log a concise, non-sensitive summary
+    try:
+        summary = {
+            'scheme': app.config.get('REDIS_SCHEME'),
+            'host': app.config.get('REDIS_URL'),
+            'port': app.config.get('REDIS_PORT'),
+            'db': app.config.get('REDIS_SPECIALIST_EXEC_DB', '0'),
+            'ssl_check_hostname': app.config.get('REDIS_SSL_CHECK_HOSTNAME'),
+            'ca_present': bool(app.config.get('REDIS_CA_CERT_PATH')),
+            'max_connections': app.config.get('REDIS_PUBSUB_MAX_CONNECTIONS'),
+            'socket_timeout': app.config.get('REDIS_PUBSUB_SOCKET_TIMEOUT'),
+            'socket_connect_timeout': app.config.get('REDIS_PUBSUB_CONNECT_TIMEOUT'),
+        }
+        app.logger.info(f"Initialized Redis pubsub pool: {summary}")
+    except Exception:
+        pass
+
+    return pool
+
+
+def get_pubsub_client(app: Flask) -> redis.Redis:
+    """Get a Redis client bound to the dedicated pubsub pool."""
+    pool = app.extensions.get('redis_pubsub_pool')
+    if pool is None:
+        pool = create_pubsub_pool(app)
+    return redis.Redis(connection_pool=pool)

config/agents/globals/RAG_AGENT/1.1.0.yaml

@@ -15,7 +15,7 @@ backstory: >
   include a salutation or closing greeting in your answer.
   {custom_backstory}
 full_model_name: "mistral.mistral-medium-latest"
-temperature: 0.5
+temperature: 0.4
 metadata:
   author: "Josako"
   date_added: "2025-01-08"

config/config.py

@@ -99,6 +99,12 @@ class Config(object):
             'ssl_check_hostname': REDIS_SSL_CHECK_HOSTNAME,
         }
 
+    # PubSub/EPT specific configuration (dedicated pool)
+    REDIS_SPECIALIST_EXEC_DB = environ.get('REDIS_SPECIALIST_EXEC_DB', '0')
+    REDIS_PUBSUB_MAX_CONNECTIONS = int(environ.get('REDIS_PUBSUB_MAX_CONNECTIONS', '200'))
+    REDIS_PUBSUB_SOCKET_TIMEOUT = float(environ.get('REDIS_PUBSUB_SOCKET_TIMEOUT', '10'))
+    REDIS_PUBSUB_CONNECT_TIMEOUT = float(environ.get('REDIS_PUBSUB_CONNECT_TIMEOUT', '3'))
+
     REDIS_PREFIXES = {
         'celery_app': 'celery:app:',
         'celery_chat': 'celery:chat:',
@@ -115,7 +121,7 @@ class Config(object):
     CELERY_RESULT_BACKEND_CHAT = f'{REDIS_BASE_URI}/0'
 
     # SSE PubSub settings
-    SPECIALIST_EXEC_PUBSUB = f'{REDIS_BASE_URI}/0'
+    SPECIALIST_EXEC_PUBSUB = f"{REDIS_BASE_URI}/{REDIS_SPECIALIST_EXEC_DB}"
 
     # eveai_model cache Redis setting
     MODEL_CACHE_URL = f'{REDIS_BASE_URI}/0'

docker/eveai_workers/Dockerfile

@@ -1,10 +1,10 @@
 FROM registry.ask-eve-ai-local.com/josakola/eveai-base:latest
 
-# Service-specific packages (ffmpeg only needed for this service)
-USER root
-RUN apt-get update && apt-get install -y --no-install-recommends \
-    ffmpeg \
-    && rm -rf /var/lib/apt/lists/*
+# Service-specific packages: ffmpeg only needed for this service - maar op dit ogenblik overbodig.
+#USER root
+#RUN apt-get update && apt-get install -y --no-install-recommends \
+#    ffmpeg \
+#    && rm -rf /var/lib/apt/lists/*
 USER appuser
 
 # Copy the service-specific source code into the container.

docker/rebuild_chat_client.sh

@@ -7,7 +7,7 @@ echo "Copying client images"
 cp -fv ../eveai_chat_client/static/assets/img/* ../nginx/static/assets/img
 
 
-dcdown eveai_chat_client nginx
+pcdown
 
 cd ../nginx
 
@@ -17,4 +17,4 @@ npm run build
 cd ../docker
 ./build_and_push_eveai.sh -b nginx
 
-dcup eveai_chat_client nginx
+pcup

documentation/Production Setup/cluster-install.md

@@ -612,6 +612,12 @@ kubectl -n eveai-staging get jobs
 kubectl -n eveai-staging logs job/<created-job-name>
 ```
 
+#### Creating volume for eveai_chat_worker's crewai storage
+
+```bash
+kubectl apply -n eveai-staging -f scaleway/manifests/base/applications/backend/eveai-chat-workers/pvc.yaml
+```
+
 #### Application Services Deployment
 Use the staging overlay to deploy apps with registry rewrite and imagePullSecrets:
 ```bash
@@ -861,3 +867,63 @@ curl https://evie-staging.askeveai.com/verify/
 
 
 
+
+
+## EveAI Chat Workers: Persistent logs storage and Celery process behavior
+
+This addendum describes how to enable persistent storage for CrewAI tuning runs under /app/logs for the eveai-chat-workers Deployment and clarifies Celery process behavior relevant to environment variables.
+
+### Celery prefork behavior and env variables
+- Pool: prefork (default). Each worker process (child) handles multiple tasks sequentially.
+- Implication: any environment variable changed inside a child process persists for subsequent tasks handled by that same child, until it is changed again or the process is recycled.
+- Our practice: set required env vars (e.g., CREWAI_STORAGE_DIR/CREWAI_STORAGE_PATH) immediately before initializing CrewAI and restore them immediately after. This prevents leakage to the next task in the same process.
+- CELERY_MAX_TASKS_PER_CHILD: the number of tasks a child will process before being recycled. Suggested starting range for heavy LLM/RAG workloads: 200–500; 1000 is acceptable if memory growth is stable. Monitor RSS and adjust.
+
+### Create and mount a PersistentVolumeClaim for /app/logs
+We persist tuning outputs under /app/logs by mounting a PVC in the worker pod.
+
+Manifests added/updated (namespace: eveai-staging):
+- scaleway/manifests/base/applications/backend/eveai-chat-workers/pvc.yaml
+- scaleway/manifests/base/applications/backend/eveai-chat-workers/deployment.yaml (volume mount added)
+
+Apply with kubectl (no Kustomize required):
+
+```bash
+# Create or update the PVC for logs
+kubectl apply -n eveai-staging -f scaleway/manifests/base/applications/backend/eveai-chat-workers/pvc.yaml
+
+# Update the Deployment to mount the PVC at /app/logs
+kubectl apply -n eveai-staging -f scaleway/manifests/base/applications/backend/eveai-chat-workers/deployment.yaml
+```
+
+Verify PVC is bound and the pod mounts the volume:
+
+```bash
+# Check PVC status
+kubectl get pvc -n eveai-staging eveai-chat-workers-logs -o wide
+
+# Inspect the pod to confirm the volume mount
+kubectl get pods -n eveai-staging -l app=eveai-chat-workers -o name
+kubectl describe pod -n eveai-staging <pod-name>
+
+# (Optional) Exec into the pod to check permissions and path
+kubectl exec -n eveai-staging -it <pod-name> -- sh -lc 'id; ls -ld /app/logs'
+```
+
+Permissions and securityContext notes:
+- The container runs as a non-root user (appuser) per Dockerfile.base. Some storage classes mount volumes owned by root. If you encounter permission issues (EACCES) writing to /app/logs:
+  - Option A: set a pod-level fsGroup so the mounted volume is group-writable by the container user.
+  - Option B: use an initContainer to chown/chmod /app/logs on the mounted volume.
+- Keep monitoring PVC usage and set alerts to avoid running out of space.
+
+Retention / cleanup recommendation:
+- For a 14-day retention, create a CronJob that runs daily to remove files older than 14 days and then delete empty directories, mounting the same PVC at /app/logs. Example command:
+
+```bash
+find /app/logs -type f -mtime +14 -print -delete; find /app/logs -type d -empty -mtime +14 -print -delete
+```
+
+Operational checks after deployment:
+1) Trigger a CrewAI tuning run; verify files appear under /app/logs and remain after pod restarts.
+2) Trigger a non-tuning run; verify temporary directories are created and cleaned up automatically.
+3) Monitor memory while varying CELERY_CONCURRENCY and CELERY_MAX_TASKS_PER_CHILD.

eveai_chat_client/static/assets/vue-components/ChatMessage.vue

@@ -6,7 +6,7 @@
                 <!-- EveAI Logo voor AI berichten - links boven, half buiten de bubbel -->
                 <img 
                     v-if="message.sender === 'ai'" 
-                    src="/static/assets/img/eveai_logo.svg"
+                    :src="staticUrl('assets/img/eveai_logo.svg')"
                     alt="EveAI" 
                     class="ai-message-logo"
                 />
@@ -105,7 +105,7 @@
                 <!-- EveAI Logo voor AI berichten - links boven, half buiten de bubbel -->
                 <img 
                     v-if="message.sender === 'ai'" 
-                    src="/static/assets/img/eveai_logo.svg"
+                    :src="staticUrl('assets/img/eveai_logo.svg')"
                     alt="EveAI" 
                     class="ai-message-logo"
                 />
@@ -124,7 +124,7 @@
                 <!-- EveAI Logo voor AI berichten - links boven, half buiten de bubbel -->
                 <img 
                     v-if="message.sender === 'ai'" 
-                    src="/static/assets/img/eveai_logo.svg"
+                    :src="staticUrl('assets/img/eveai_logo.svg')"
                     alt="EveAI" 
                     class="ai-message-logo"
                 />
@@ -166,11 +166,23 @@ export default {
             originalTexts
         );
         
+        // Helper to build environment-aware static asset URLs
+        const staticUrl = (p) => {
+            try {
+                if (typeof window !== 'undefined' && typeof window.staticUrl === 'function') {
+                    return window.staticUrl(p);
+                }
+            } catch (e) {}
+            const base = '/static/';
+            return base + String(p || '').replace(/^\/+/, '');
+        };
+        
         return {
             messageTexts: translations,
             translationLoading: isLoading,
             translationError: error,
-            currentLanguage
+            currentLanguage,
+            staticUrl
         };
     },
     props: {

eveai_chat_client/static/assets/vue-components/ProgressTracker.vue

@@ -14,7 +14,7 @@
             <div class="progress-title">
                 <!-- Evie working animatie tijdens processing -->
                 <img v-if="isProcessing" 
-                     src="/static/assets/img/evie_working.webp" 
+                     :src="staticUrl('assets/img/evie_working.webp')" 
                      alt="Bezig met verwerken..." 
                      class="progress-icon working-animation">
                 
@@ -55,7 +55,7 @@
     
     <!-- Alleen Evie animatie voor "No Information" tijdens processing -->
     <div v-else-if="shouldShowProgressIconOnly" class="progress-icon-only">
-        <img src="/static/assets/img/evie_working.webp" 
+        <img :src="staticUrl('assets/img/evie_working.webp')" 
              alt="Bezig met verwerken..." 
              class="working-animation-only">
     </div>
@@ -72,6 +72,16 @@ import { useComponentTranslations } from '../js/services/LanguageProvider.js';
 export default {
     name: 'ProgressTracker',
     setup() {
+        // Helper to build environment-aware static asset URLs
+        const staticUrl = (p) => {
+            try {
+                if (typeof window !== 'undefined' && typeof window.staticUrl === 'function') {
+                    return window.staticUrl(p);
+                }
+            } catch (e) {}
+            const base = '/static/';
+            return base + String(p || '').replace(/^\/+/, '');
+        };
         // Define original English texts (base language for developers)
         const originalTexts = {
             error: 'Error while processing',
@@ -92,7 +102,8 @@ export default {
             statusTexts: translations,
             translationLoading: isLoading,
             translationError: error,
-            currentLanguage
+            currentLanguage,
+            staticUrl
         };
     },
     props: {

eveai_chat_client/templates/chat.html

@@ -28,7 +28,9 @@
         tenantMake: {
             name: "{{ tenant_make.name|default('EveAI') }}",
             logo_url: "{{ tenant_make.logo_url|default('') }}"
-        }
+        },
+        // Environment-aware static base provided by Flask's overridden url_for
+        staticBase: '{{ static_url }}'
     };
 
     // Debug info om te controleren of chatConfig correct is ingesteld

eveai_chat_client/templates/scripts.html

@@ -1,6 +1,30 @@
 <!-- Material Icons voor icoontjes in de interface -->
 <link rel="stylesheet" href="https://fonts.googleapis.com/css2?family=Material+Symbols+Outlined:opsz,wght,FILL,GRAD@24,400,0,0">
 
+<!-- Define a robust global staticUrl helper BEFORE loading the chat client bundle -->
+<script>
+    (function(){
+        if (typeof window === 'undefined') return;
+        // If already defined (e.g., by chat.html head), reuse it and expose globally
+        if (typeof window.staticUrl === 'function') {
+            try { globalThis.staticUrl = window.staticUrl; } catch (e) {}
+        } else {
+            // Prefer runtime chatConfig.staticBase; else fallback to server-provided base or default
+            var serverStaticBase = '{{ static_url|default("") }}' || '';
+            if (!serverStaticBase) { serverStaticBase = '{{ url_for("static", filename="") }}'; }
+            var base = (window.chatConfig && window.chatConfig.staticBase) ? window.chatConfig.staticBase : (serverStaticBase || '/static/');
+            var normalizedBase = String(base).replace(/\/+$/, '/');
+            window.staticUrl = function(path) {
+                if (!path) return normalizedBase;
+                return normalizedBase + String(path).replace(/^\/+/, '');
+            };
+            try { globalThis.staticUrl = window.staticUrl; } catch (e) {}
+        }
+        // Optional lightweight debug
+        // console.debug('[scripts.html] staticUrl defined:', typeof window.staticUrl);
+    })();
+</script>
+
 <!-- Chat client JS - bundled met alle componenten en ES modules -->
 <script src="{{url_for('static', filename='dist/chat-client.js')}}"></script>
 <script>

eveai_chat_client/views/chat_views.py

@@ -1,6 +1,7 @@
 import json
 import uuid
-from flask import Blueprint, render_template, request, session, current_app, jsonify, Response, stream_with_context
+from flask import Blueprint, render_template, request, session, current_app, jsonify, Response, stream_with_context, \
+    url_for
 from sqlalchemy.exc import SQLAlchemyError
 
 from common.extensions import db, content_manager
@@ -110,6 +111,11 @@ def chat(magic_link_code):
         if isinstance(specialist_config, str):
             specialist_config = json.loads(specialist_config)
 
+        static_url = current_app.config.get('STATIC_URL')
+        current_app.logger.debug(f"STATIC_URL: {static_url}")
+        if not static_url:
+            static_url = url_for('static', filename='')
+
         return render_template('chat.html',
                                tenant=tenant,
                                tenant_make=tenant_make,
@@ -117,7 +123,8 @@ def chat(magic_link_code):
                                customisation=customisation,
                                messages=[],
                                settings=settings,
-                               config=current_app.config
+                               config=current_app.config,
+                               static_url=static_url
                                )
 
     except Exception as e:

eveai_chat_workers/__init__.py

@@ -32,6 +32,13 @@ def create_app(config_file=None):
 
     register_cache_handlers(app)
 
+    # Initialize dedicated Redis pubsub pool for ExecutionProgressTracker
+    try:
+        from common.utils.redis_pubsub_pool import create_pubsub_pool
+        create_pubsub_pool(app)
+    except Exception as e:
+        app.logger.error(f"Failed to initialize Redis pubsub pool: {e}")
+
     from . import specialists, retrievers
 
     celery = make_celery(app.name, app.config)

eveai_chat_workers/specialists/crewai_base_classes.py

@@ -1,5 +1,7 @@
 import asyncio
 import json
+import os
+import tempfile
 import time
 from abc import abstractmethod
 
@@ -8,7 +10,7 @@ from crewai.agents.parser import AgentAction, AgentFinish
 from crewai.crews import CrewOutput
 from crewai.tools import BaseTool
 from flask import current_app
-from pydantic import BaseModel, create_model, Field, ConfigDict
+from pydantic import BaseModel, create_model, Field, ConfigDict, PrivateAttr
 from typing import Dict, Type, get_type_hints, Optional, List, Any, Callable
 
 from common.utils.business_event_context import current_event
@@ -74,12 +76,20 @@ class EveAICrewAICrew(Crew):
     specialist: Any = Field(default=None, exclude=True)
     name: str = Field(default=None, exclude=True)
     model_config = ConfigDict(arbitrary_types_allowed=True)
+    _tmp_dir = PrivateAttr(default=None)
 
     def __init__(self, specialist, name: str, **kwargs):
         if specialist.tuning:
-            kwargs['output_log_file'] = f"/app/logs/crew_{name}.txt"
+            base_dir = f"/app/logs/crew_{name}/{specialist.task_id}"
+        else:
+            self._tmp_dir = tempfile.TemporaryDirectory(prefix="crewai-")
+            base_dir = self._tmp_dir.name
+        os.environ["CREWAI_STORAGE_DIR"] = base_dir
+        os.environ["CREWAI_STORAGE_PATH"] = base_dir
+        kwargs['output_log_file'] = f"{base_dir}/specialist_log.json"
 
         super().__init__(**kwargs)
+
         self.specialist = specialist
         self.name = name
 

scaleway/manifests/base/applications/backend/eveai-chat-workers/deployment.yaml

@@ -54,6 +54,9 @@ spec:
               name: eveai-secrets
               key: PUSH_GATEWAY_PORT
               optional: true
+        volumeMounts:
+        - name: logs-volume
+          mountPath: /app/logs
         resources:
           requests:
             cpu: "500m"
@@ -61,3 +64,8 @@ spec:
           limits:
             cpu: "2"
             memory: "3Gi"
+
+      volumes:
+      - name: logs-volume
+        persistentVolumeClaim:
+          claimName: eveai-chat-workers-logs