Auto-generated

This page is auto-generated from the Helm chart source. Do not edit directly. Regenerate with make docs-helm-charts.

supervisor-agent

A Helm chart for the Supervisor Agent

Version	0.2.38
Type	application

Quick Start

# Add and install the chart
helm install supervisor-agent oci://ghcr.io/cnoe-io/charts/supervisor-agent --version 0.2.38

# Upgrade an existing release
helm upgrade supervisor-agent oci://ghcr.io/cnoe-io/charts/supervisor-agent --version 0.2.38

Customizing Values

Override default values using --set flags or a custom values file:

# Override individual values
helm install supervisor-agent oci://ghcr.io/cnoe-io/charts/supervisor-agent --version 0.2.38 \
  --set replicaCount=2

# Use a custom values file
helm install supervisor-agent oci://ghcr.io/cnoe-io/charts/supervisor-agent --version 0.2.38 \
  -f custom-values.yaml

# Show all configurable values
helm show values oci://ghcr.io/cnoe-io/charts/supervisor-agent --version 0.2.38

Reading the Values Table

Column	Meaning
Key	Dot-separated path into values.yaml (e.g. image.repository)
Type	Go/Helm data type (string, int, bool, object, list)
Default	Value used when not overridden
Description	What the parameter controls

Values

Key	Type	Default	Description
affinity	object	{}	Pod affinity rules
autoscaling.enabled	bool	false	Enable Horizontal Pod Autoscaler
autoscaling.maxReplicas	int	100	HPA maximum replicas
autoscaling.minReplicas	int	1	HPA minimum replicas
autoscaling.targetCPUUtilizationPercentage	int	80	HPA target CPU utilization percentage
checkpointPersistence.mongodb.existingSecret.key	string	""	Key inside the secret
checkpointPersistence.mongodb.existingSecret.name	string	""	Name of a Kubernetes Secret containing the MongoDB URI
checkpointPersistence.mongodb.uri	string	""	MongoDB connection URI
checkpointPersistence.postgres.dsn	string	""	PostgreSQL DSN
checkpointPersistence.postgres.existingSecret.key	string	""	Key inside the secret
checkpointPersistence.postgres.existingSecret.name	string	""	Name of a Kubernetes Secret containing the Postgres DSN
checkpointPersistence.redis.autoDiscoverService	string	""	In-cluster Redis service name to auto-build the URL from
checkpointPersistence.redis.dbIndex	int	0	Redis DB index
checkpointPersistence.redis.existingSecret.key	string	""	Key inside the secret
checkpointPersistence.redis.existingSecret.name	string	""	Name of a Kubernetes Secret containing the Redis URL
checkpointPersistence.redis.url	string	""	Redis URL. Requires Redis Stack with RedisJSON + RediSearch modules
checkpointPersistence.ttlMinutes	int	0	TTL for checkpoints in minutes. 0 disables expiry
checkpointPersistence.type	string	"memory"	Checkpoint backend. One of: memory (in-process default), redis, postgres, mongodb
env	object	{"EXTERNAL_URL":"http://localhost:8000"}	Non-sensitive environment variables injected into the supervisor container
externalSecrets.enabled	bool	false	Enable External Secrets Operator integration
externalSecrets.secretNames	list	[]	List of existing ESO secret names to reference
fullnameOverride	string	""	Override the full release name
image.args	list	["platform-engineer"]	Container args
image.pullPolicy	string	"Always"	Image pull policy
image.repository	string	"ghcr.io/cnoe-io/ai-platform-engineering"	Container image repository
image.tag	string	""	Image tag. Defaults to chart appVersion when not set
imagePullSecrets	list	[]	Image pull secrets for private registries
ingress.annotations	object	{}	Ingress annotations
ingress.className	string	""	Ingress class name
ingress.enabled	bool	false	Enable ingress
ingress.hosts	list	[]	Ingress host rules
ingress.tls	list	[]	Ingress TLS configuration
livenessProbe	object	{"failureThreshold":3,"initialDelaySeconds":30,"periodSeconds":10,"tcpSocket":{"port":"http"},"timeoutSeconds":5}	Liveness probe configuration
llmSecrets.create	bool	true	Create the LLM secret. Set false if the secret already exists in the cluster
llmSecrets.data	object	{}	Key/value pairs for the LLM secret
llmSecrets.externalSecrets.data	list	[]	ESO data mappings
llmSecrets.externalSecrets.enabled	bool	false	Use External Secrets Operator for the LLM secret
llmSecrets.externalSecrets.secretStoreRef.kind	string	"ClusterSecretStore"	ESO SecretStore kind
llmSecrets.externalSecrets.secretStoreRef.name	string	""	ESO SecretStore name
llmSecrets.secretName	string	"llm-secret"	Name of the LLM secret
memoryPersistence.embeddings.dims	string	""	Embedding dimensions. Auto-detected for known models when empty
memoryPersistence.embeddings.model	string	""	Embedding model name (e.g. text-embedding-3-small)
memoryPersistence.embeddings.provider	string	""	Embedding provider for semantic memory search (openai, azure-openai, litellm)
memoryPersistence.enableFactExtraction	bool	false	Automatically extract facts from every conversation turn and store them for cross-thread recall
memoryPersistence.maxMemories	int	50	Maximum number of memory facts returned to the LLM per cross-thread context query
memoryPersistence.maxSummaries	int	10	Maximum number of conversation summaries returned to the LLM per cross-thread context query
memoryPersistence.mongodb.existingSecret.key	string	""	Key inside the secret (e.g. MONGODB_URI)
memoryPersistence.mongodb.existingSecret.name	string	""	Name of a Kubernetes Secret containing the MongoDB URI
memoryPersistence.mongodb.uri	string	""	MongoDB connection URI (e.g. mongodb://host:27017)
memoryPersistence.postgres.dsn	string	""	PostgreSQL DSN (e.g. postgresql://user:pass@host:5432/db). Prefer existingSecret for credentials
memoryPersistence.postgres.existingSecret.key	string	""	Key inside the secret (e.g. DATABASE_URL)
memoryPersistence.postgres.existingSecret.name	string	""	Name of a Kubernetes Secret containing the Postgres DSN
memoryPersistence.redis.autoDiscoverService	string	""	In-cluster Redis service name to auto-build the URL from (redis://release-service:6379/dbIndex``). Common values: langgraph-redis, rag-redis
memoryPersistence.redis.dbIndex	int	0	Redis DB index. Use 1+ when sharing with RAG Redis (db 0 is RAG data)
memoryPersistence.redis.existingSecret.key	string	""	Key inside the secret (e.g. REDIS_URL)
memoryPersistence.redis.existingSecret.name	string	""	Name of a Kubernetes Secret containing the Redis URL
memoryPersistence.redis.keyPrefix	string	""	Key namespace prefix for shared Redis instances. Prevents key collisions across deployments
memoryPersistence.redis.url	string	""	Redis URL (e.g. redis://host:6379/0). Takes precedence over autoDiscoverService
memoryPersistence.ttlMinutes	int	10080	TTL for stored memory items in minutes. 0 disables expiry
memoryPersistence.type	string	"memory"	Memory store backend. One of: memory (in-process default), redis, postgres, mongodb
metrics.enabled	bool	true	Expose /metrics endpoint for Prometheus scraping
multiAgentConfig.agents	list	[]	Explicit agent list. Auto-discovered when empty
multiAgentConfig.port	string	"8000"	Port advertised to peer agents
multiAgentConfig.protocol	string	"a2a"	Agent-to-agent protocol (a2a or slim)
nameOverride	string	""	Override the chart name
nodeSelector	object	{}	Node selector labels
podAnnotations	object	{}	Extra annotations added to the pod
podLabels	object	{}	Extra labels added to the pod
podSecurityContext	object	{}	Pod-level security context
readinessProbe	object	{"failureThreshold":3,"initialDelaySeconds":5,"periodSeconds":5,"tcpSocket":{"port":"http"},"timeoutSeconds":3}	Readiness probe configuration
replicaCount	int	1	Number of replicas
resources	object	{}	CPU/memory requests and limits
revisionHistoryLimit	int	3	Number of old ReplicaSets to retain for rollback
securityContext	object	{}	Container-level security context
service.port	int	8000	Service port
service.type	string	"ClusterIP"	Kubernetes service type
serviceAccount.annotations	object	{}	Annotations to add to the service account
serviceAccount.automount	bool	true	Automount the service account API token
serviceAccount.create	bool	true	Create a service account
serviceAccount.name	string	""	Override the service account name. Generated from fullname template when empty
slim.enabled	bool	false	Enable SLIM transport (standalone fallback; in production provided by global.slim)
slim.endpoint	string	"http://ai-platform-engineering-slim:46357"	SLIM endpoint URL
slim.transport	string	"slim"	Transport label passed to agents
tolerations	list	[]	Pod tolerations
volumeMounts	list	[]	Extra volume mounts
volumes	list	[]	Extra volumes
vpa	object	{"controlledResources":["cpu","memory"],"enabled":false,"maxAllowed":{},"minAllowed":{"cpu":"50m","memory":"128Mi"},"updateMode":"InPlaceOrRecreate"}	Vertical Pod Autoscaler - applies resource recommendations automatically (updateMode: "InPlaceOrRecreate"). Requires VPA to be installed in the cluster before deploying with vpa.enabled=true. Do not enable alongside autoscaling.enabled - VPA conflicts with HPA on both CPU-based and memory-based scaling. Setting initial resource requests (see resources: above) improves the quality of early VPA recommendations.
vpa.controlledResources	list	["cpu","memory"]	Resources VPA should observe and recommend for
vpa.enabled	bool	false	Enable VPA
vpa.maxAllowed	object	{}	Maximum bounds VPA will recommend (empty = no cap)
vpa.minAllowed	object	{"cpu":"50m","memory":"128Mi"}	Minimum bounds VPA will recommend
vpa.updateMode	string	"InPlaceOrRecreate"	Use "Off" for recommendations only (no restarts), or "Recreate" on older clusters.