Status Report

The platform’s health now, in one report: everything that the status services read, and what looks wrong.

platform_status() reads the environment’s pods and the cluster’s nodes, with their CPU and memory use, the namespace’s warning events (cluster), the dashboard’s health checks, the Celery queues and the database’s load (platform), and the status pages of the AI providers and of AWS (feeds). Its report starts with its observations: what looks wrong, in plain sentences, from observations().
eks_status() reads the cluster as the cloud provider sees it: the control plane, the node groups, the add-ons and their available updates, and the provider’s insights, from get_kubernetes_cluster_health(), and the recent AWS events of the platform’s region.

The MCPServer’s staff-only tools platform_status and eks_status (smarter.apps.mcpserver.protocol.tools.status) format these reports as Markdown. Nothing adds cloud monitoring costs: the data comes from Kubernetes, metrics-server, the platform’s own services, the EKS api, and public status pages.

Note

Experimental. The MCPServer was designed and coded by Claude Code (Anthropic’s Claude Opus 5.5), with Lawrence McDaniel as co-author. It is experimental, and will be documented.

smarter.apps.mcpserver.services.status.report.DATABASE_BUSY_PERCENT = 80: The share of the database’s connection limit in use at which the database is an observation.

class smarter.apps.mcpserver.services.status.report.EksStatus(health, region, aws_events, unread_feeds)[source]

Bases: object

The cluster, as the cloud provider sees it, as eks_status() returns it.

__init__(health, region, aws_events, unread_feeds)

aws_events: list[AwsEvent]: The recent AWS events of EKS and EC2 in the region.

health: ClusterHealth: The cluster’s control plane, node groups, add-ons and insights.

region: str: The platform’s AWS region, from platform_region().

unread_feeds: list[str]: The AWS services whose status feed could not be read.

smarter.apps.mcpserver.services.status.report.HIGH_USE_PERCENT = 90

The use, as a percentage of what is allowed, at which a pod’s memory, or a node’s CPU or memory, is an observation.

A pod above its memory limit is killed (OOMKilled), and a node that is full cannot take new pods.

class smarter.apps.mcpserver.services.status.report.PlatformStatus(namespace, region, snapshot, checks, depths, database, providers, aws_events, unread_feeds, observations)[source]

Bases: object

The platform’s health now, as platform_status() returns it.

__init__(namespace, region, snapshot, checks, depths, database, providers, aws_events, unread_feeds, observations)

aws_events: list[AwsEvent]: The recent AWS events of EKS and EC2 in the region.

checks: list[tuple[str, bool]]

each one’s name, and whether it passed.

Type:: The dashboard’s health checks

database: DatabaseLoad | None: The database’s load, or None if it could not be read.

depths: dict[str, int | None]: The number of tasks waiting in each Celery queue, or None for a queue that could not be read.

namespace: str: The environment’s Kubernetes namespace.

observations: list[str]: What looks wrong, in plain sentences, from observations().

providers: list[ProviderStatus]: The AI providers’ status pages.

region: str: The platform’s AWS region, from platform_region().

snapshot: ClusterSnapshot: The environment’s pods, the cluster’s nodes, and the namespace’s warning events.

unread_feeds: list[str]: The AWS services whose status feed could not be read.

smarter.apps.mcpserver.services.status.report.QUEUE_BACKLOG = 50

the workers are behind.

Type:: The number of tasks waiting in a Celery queue at which the queue is an observation

smarter.apps.mcpserver.services.status.report.eks_status()[source]

Read the cluster’s health from the cloud provider, and the recent AWS events of its region.

Side effects: read-only calls to the EKS api (DescribeCluster, ListNodegroups, DescribeNodegroup, ListAddons, DescribeAddon, DescribeAddonVersions and ListInsights), and up to two HTTPS GETs of AWS status feeds, which are cached for a minute.

Return type:

EksStatus

Returns:

The cluster’s health, and the region’s AWS events.

Raises:

InfrastructureNotReadyError – If the cloud provider is not ready.
SmarterInfrastructureError – If the cluster cannot be described.

smarter.apps.mcpserver.services.status.report.node_observations(nodes)[source]

Describe what is wrong with the cluster’s nodes.

A node is an observation when it is not ready, reports a pressure condition, e.g. MemoryPressure, is cordoned, or uses at least HIGH_USE_PERCENT of its CPU or memory.

Parameters:: nodes (list[NodeStatus]) – The nodes, from cluster_snapshot().
Return type:: list[str]
Returns:: One sentence per problem, or an empty list if there is none.

smarter.apps.mcpserver.services.status.report.observations(snapshot, checks, depths, database, providers, aws_events)[source]

Describe everything that looks wrong, for the top of the platform’s status report.

It gathers the observations of the pods, the nodes, the warning events, the backing services and the upstream status pages, and the parts of the cluster that could not be read.

Parameters:

snapshot (ClusterSnapshot) – The pods, nodes and warning events, from cluster_snapshot().
checks (list[tuple[str, bool]]) – The dashboard’s health checks.
depths (dict[str, int | None]) – The Celery queues’ depths.
database (DatabaseLoad | None) – The database’s load, or None if it could not be read.
providers (list[ProviderStatus]) – The AI providers’ status.
aws_events (list[AwsEvent]) – The recent AWS events.

Return type:

list[str]

Returns:

One sentence per problem, or an empty list if nothing looks wrong.

smarter.apps.mcpserver.services.status.report.platform_region()[source]

Return the platform’s AWS region, for the AWS status feeds and the report.

smarter_settings.aws_region is None when no region is configured. It is then reported, and put in the feeds’ urls, as None, as the tools have always done. No AWS feed exists for a region of that name.

Return type:: str
Returns:: The region, e.g. us-east-2, or "None" if none is configured.

smarter.apps.mcpserver.services.status.report.platform_status(minutes)[source]

Read the platform’s health now, and what looks wrong.

Side effects: five kubectl get calls, a SELECT 1 and two SHOW queries on the database, a value written to and deleted from the cache, a ping of the Celery workers, a connection to the Celery broker, and up to five HTTPS GETs of public status pages, which are cached for a minute. Nothing in the cluster, the database or the queues is changed.

Parameters:: minutes (int) – How far back to read Kubernetes warning events.
Return type:: PlatformStatus
Returns:: Everything that was read, and the observations. A part that cannot be read is named in the observations, and the others are still read.

smarter.apps.mcpserver.services.status.report.pod_observations(pods)[source]

Describe what is wrong with the environment’s pods.

A pod is an observation when it is not ready, reports a problem, e.g. CrashLoopBackOff, has restarted, or uses at least HIGH_USE_PERCENT of its memory limit.

Parameters:: pods (list[PodStatus]) – The pods, from cluster_snapshot().
Return type:: list[str]
Returns:: One sentence per problem, or an empty list if there is none.

smarter.apps.mcpserver.services.status.report.read_database()[source]

Read the database’s load, or report that it could not be read.

Return type:: DatabaseLoad | None
Returns:: The load, or None if the database could not be queried. The failure is logged, and becomes an observation.

smarter.apps.mcpserver.services.status.report.service_observations(checks, depths, database)[source]

Describe what is wrong with the platform’s backing services.

A failed health check is an observation, and so is a Celery queue that could not be read or has at least QUEUE_BACKLOG waiting tasks, and a database with at least DATABASE_BUSY_PERCENT of its connections in use.

Parameters:

checks (list[tuple[str, bool]]) – The dashboard’s health checks, from health_checks().
depths (dict[str, int | None]) – The Celery queues’ depths, from queue_depths().
database (DatabaseLoad | None) – The database’s load, or None if it could not be read.

Return type:

list[str]

Returns:

One sentence per problem, or an empty list if there is none.

smarter.apps.mcpserver.services.status.report.upstream_observations(providers, aws_events)[source]

Describe the problems that the AI providers and AWS report on their status pages.

A provider whose page could not be read is not an observation: its status is unknown, not bad, and the report’s upstream section says so.

Parameters:

providers (list[ProviderStatus]) – The AI providers’ status, from provider_statuses().
aws_events (list[AwsEvent]) – The recent AWS events, from aws_service_events().

Return type:

list[str]

Returns:

One sentence per problem, or an empty list if there is none.

smarter.apps.mcpserver.services.status.report.utc(value)[source]

Format a time for a report.

Parameters:: value (datetime | None) – The time, or None if it is unknown.
Return type:: str
Returns:: The time in UTC, e.g. 2026-11-02 14:05:00 UTC, or ? if it is unknown.