Version: 1.0

Tool Reference

Complete reference for all 49 tools exposed by the KubeOpera MCP Server.

Read-Only Data Tools

These tools query live data from KubeOpera services. All are marked ReadOnlyHint: true and OpenWorld: true.

get_cluster_health

Fetches the current cluster health score, node status summary, pod counts, and API server latency.

Source service: k8s-monitor (port 8085) Endpoint: GET /api/health

Parameters: None

Example response:

{
  "health_score": 87,
  "status": "healthy",
  "nodes": {
    "total": 6,
    "ready": 6,
    "not_ready": 0
  },
  "pods": {
    "total": 142,
    "running": 138,
    "pending": 2,
    "failed": 2,
    "succeeded": 0
  },
  "api_server_latency_ms": 12,
  "checked_at": "2026-04-18T10:00:00Z"
}

get_cluster_cost

Fetches current cloud cost breakdown by namespace, workload, and resource type.

Source service: k8s-monitor (port 8085) Endpoint: GET /api/cost

Parameters: None

Example response:

{
  "total_monthly_usd": 4820.50,
  "by_namespace": [
    { "namespace": "production", "monthly_usd": 3200.00, "cpu_pct": 66, "memory_pct": 70 },
    { "namespace": "staging", "monthly_usd": 980.00, "cpu_pct": 20, "memory_pct": 18 }
  ],
  "by_resource": {
    "compute_usd": 3100.00,
    "memory_usd": 1200.00,
    "storage_usd": 520.50
  }
}

get_optimization_report

Returns over-provisioned and idle workload recommendations with estimated monthly savings.

Source service: k8s-monitor (port 8085) Endpoint: GET /api/optimizer

Parameters:

Name	Type	Required	Description
`namespace`	string	No	Filter recommendations to a specific namespace
`view`	string	No	`summary` (default) or `detailed`

Example response:

{
  "total_savings_usd": 640.00,
  "recommendations": [
    {
      "workload": "payments-api",
      "namespace": "production",
      "type": "over_provisioned",
      "current_cpu_request": "2000m",
      "recommended_cpu_request": "500m",
      "savings_usd": 220.00
    }
  ]
}

get_pod_metrics

Returns per-pod CPU and memory usage for all pods or a specific namespace.

Source service: k8s-monitor (port 8085) Endpoint: GET /api/metrics/pods

Parameters:

Name	Type	Required	Description
`namespace`	string	No	Filter to a specific namespace

Example response:

{
  "pods": [
    {
      "name": "payments-api-6d8f9b-xkj2p",
      "namespace": "production",
      "cpu_usage_millicores": 420,
      "memory_usage_mb": 312,
      "cpu_limit_millicores": 2000,
      "memory_limit_mb": 512
    }
  ],
  "collected_at": "2026-04-18T10:00:00Z"
}

get_node_metrics

Returns per-node CPU, memory, and disk metrics for all cluster nodes.

Source service: k8s-monitor (port 8085) Endpoint: GET /api/metrics/nodes

Parameters: None

Example response:

{
  "nodes": [
    {
      "name": "ip-10-0-1-45.ec2.internal",
      "cpu_usage_pct": 62,
      "memory_usage_pct": 74,
      "disk_usage_pct": 38,
      "ready": true,
      "roles": ["worker"]
    }
  ]
}

get_security_posture

Returns compliance score, vulnerability counts, and RBAC findings.

Source service: security-api (port 8086) Endpoint: GET /api/v1/posture/summary

Parameters:

Name	Type	Required	Description
`cluster_id`	string	No	Filter to a specific cluster

Example response:

{
  "compliance_score": 78,
  "vulnerabilities": {
    "critical": 2,
    "high": 8,
    "medium": 24,
    "low": 51
  },
  "rbac_findings": 5,
  "privileged_pods": 1,
  "last_scan_at": "2026-04-18T09:30:00Z"
}

get_pipeline_status

Returns CI/CD run history, success rate, and failed stages.

Source service: cicd-gateway (port 8087) Endpoint: GET /api/v1/clusters/{cluster_id}/runs/summary

Parameters:

Name	Type	Required	Description
`cluster_id`	string	No	Filter to a specific cluster
`limit`	integer	No	Number of recent runs (default: 20)

Example response:

{
  "success_rate_pct": 92,
  "avg_duration_secs": 187,
  "total_runs_24h": 38,
  "failed_runs": [
    {
      "id": "run-abc123",
      "pipeline": "payments-api",
      "branch": "main",
      "failed_stage": "integration-tests",
      "started_at": "2026-04-18T08:15:00Z"
    }
  ]
}

get_cluster_list

Returns all registered clusters with health score, node count, and resource data.

Source service: kubeopera-api (port 8080) Endpoint: GET /api/v1/clusters

Parameters: None

Example response:

{
  "clusters": [
    {
      "id": "prod-us-east",
      "name": "Production US East",
      "provider": "aws",
      "region": "us-east-1",
      "node_count": 6,
      "health_score": 87,
      "status": "running"
    }
  ],
  "total": 3
}

get_anomaly_events

Returns anomaly events with Z-score, severity, and auto-remediation status.

Source service: anomaly-detector (port 8088) Endpoint: GET /api/v1/anomalies

Parameters:

Name	Type	Required	Description
`cluster_id`	string	No	Filter by cluster
`severity`	string	No	`low`, `medium`, `high`, or `critical`
`limit`	integer	No	Number of results (default: 20)

Example response:

{
  "anomalies": [
    {
      "id": "evt-7f3d",
      "cluster_id": "prod-us-east",
      "metric": "cpu_usage_pct",
      "value": 94.2,
      "baseline": 45.1,
      "z_score": 3.8,
      "severity": "critical",
      "namespace": "production",
      "resource": "payments-api",
      "healing_id": "heal-abc",
      "acknowledged": false,
      "detected_at": "2026-04-18T09:55:00Z"
    }
  ]
}

get_scaling_forecasts

Returns Holt-Winters load forecasts and proactive scaling recommendations.

Source service: predictive-scaler (port 8089) Endpoint: GET /api/v1/scaling-decisions and GET /api/v1/forecasts

Parameters:

Name	Type	Required	Description
`cluster_id`	string	No	Filter by cluster
`status`	string	No	`pending`, `approved`, `applied`, or `rejected`

Example response:

{
  "forecasts": [
    {
      "workload": "payments-api",
      "namespace": "production",
      "metric": "cpu_usage_pct",
      "horizon_mins": 60,
      "predicted_peak": 87.4,
      "confidence": 0.91
    }
  ],
  "scaling_decisions": [
    {
      "id": "dec-9a2f",
      "workload": "payments-api",
      "current_replicas": 3,
      "recommended_replicas": 5,
      "reason": "CPU forecast exceeds 80% in 45 minutes",
      "status": "pending"
    }
  ]
}

get_incidents

Returns active and recent incidents with MTTR and runbook status.

Source service: incident-manager (port 8090) Endpoint: GET /api/v1/incidents and GET /api/v1/incidents/metrics

Parameters:

Name	Type	Required	Description
`cluster_id`	string	No	Filter by cluster
`status`	string	No	`open`, `investigating`, `mitigating`, or `resolved`

Example response:

{
  "incidents": [
    {
      "id": "inc-f4a1",
      "title": "High CPU usage on payments-api",
      "severity": "high",
      "status": "investigating",
      "cluster_id": "prod-us-east",
      "namespace": "production",
      "mttr_secs": null,
      "runbook_id": "rb-cpu-scale",
      "opened_at": "2026-04-18T09:56:00Z"
    }
  ],
  "metrics": {
    "open_count": 2,
    "avg_mttr_secs": 840,
    "sla_breach_pct": 0
  }
}

get_multi_cluster_overview

Returns aggregated health, cost, and incident count across all registered clusters.

Source service: kubeopera-api (port 8080) + k8s-monitor (port 8085) — parallel fetch Endpoint: Multiple

Parameters: None

Example response:

{
  "total_clusters": 3,
  "total_nodes": 18,
  "avg_health_score": 83,
  "total_cost_monthly_usd": 14200.00,
  "open_incidents": 2,
  "clusters": [
    {
      "id": "prod-us-east",
      "health_score": 87,
      "node_count": 6,
      "cost_monthly_usd": 4820.50,
      "open_incidents": 1
    }
  ]
}

get_agent_telemetry

Returns recent telemetry snapshots from the observability agent.

Source service: observability-agent-srv (port 8092) Endpoint: GET /api/v1/telemetry

Parameters:

Name	Type	Required	Description
`cluster_id`	string	No	Filter by cluster
`limit`	integer	No	Number of snapshots (default: 10)

get_analysis_results

Returns statistical analysis results with risk scores and anomaly decisions.

Source service: analysis-agent-srv (port 8093) Endpoint: GET /api/v1/results

Parameters:

Name	Type	Required	Description
`cluster_id`	string	No	Filter by cluster
`limit`	integer	No	Number of results (default: 10)

get_action_log

Returns the automated Kubernetes action log with outcomes.

Source service: action-agent-srv (port 8094) Endpoint: GET /api/v1/actions

Parameters:

Name	Type	Required	Description
`cluster_id`	string	No	Filter by cluster
`limit`	integer	No	Number of actions (default: 20)

get_feedback_outcomes

Returns feedback signals showing whether automated actions improved cluster state.

Source service: feedback-agent-srv (port 8095) Endpoint: GET /api/v1/feedback

Parameters:

Name	Type	Required	Description
`cluster_id`	string	No	Filter by cluster
`limit`	integer	No	Number of signals (default: 20)

get_recommendations

Returns AI-generated recommendations from the recommendation agent.

Source service: recommendation-agent-srv (port 8096) Endpoint: GET /api/v1/recommendations

Parameters:

Name	Type	Required	Description
`cluster_id`	string	No	Filter by cluster
`limit`	integer	No	Number of recommendations (default: 10)

Example response:

{
  "recommendations": [
    {
      "id": "rec-b3c2",
      "cluster_id": "prod-us-east",
      "category": "performance",
      "priority": "high",
      "title": "Scale payments-api deployment",
      "description": "CPU utilisation exceeded 85% for 4 consecutive cycles (Z-score 3.1).",
      "impact": "Request latency will increase. Risk of OOMKill under current load.",
      "action": "Increase replicas from 3 to 5. Review HPA maxReplicas (currently 4).",
      "risk_score": 72.5,
      "generated_at": "2026-04-18T09:58:00Z"
    }
  ]
}

get_alert_rules

Returns anomaly alert rules with metric, threshold, severity, and configured auto-heal action.

Source service: anomaly-detector (port 8088) Endpoint: GET /api/v1/alert-rules

Parameters:

Name	Type	Required	Description
`cluster_id`	string	No	Filter by cluster

execute_runbook

Triggers execution of an automated remediation runbook against an incident. Execution is asynchronous — use get_runbook_execution to track progress.

Source service: incident-manager (port 8090) Endpoint: POST /api/v1/incidents/{id}/runbooks/{rbId}/execute

Parameters:

Name	Type	Required	Description
`incident_id`	string	Yes	Incident to execute against
`runbook_id`	string	Yes	Runbook to run
`started_by`	string	No	Operator identifier (default: `mcp-agent`)

Returns: { id, status: "pending", incident_id, runbook_id, started_by, started_at }

get_runbook_execution

Returns the current status and per-step results of a runbook execution.

Source service: incident-manager (port 8090) Endpoint: GET /api/v1/executions/{execId}

Parameters:

Name	Type	Required	Description
`execution_id`	string	Yes	Execution ID from `execute_runbook`

generate_post_incident_report

Generates an AI-powered post-incident report for a resolved incident using Claude Haiku. Returns structured summary, root_cause, impact, and action_items.

Source service: incident-manager (port 8090) Endpoint: POST /api/v1/incidents/{id}/pir

Parameters:

Name	Type	Required	Description
`incident_id`	string	Yes	Incident to generate report for

get_node_pools

Lists Karpenter NodePools with resource limits, consolidation policy, and node/claim counts.

Source service: nodes-manager (port 8098) Endpoint: GET /api/v1/nodepools

Parameters: None

get_node_claims

Lists NodeClaims for a Karpenter NodePool showing phase (Pending/Bound/Disrupted), instance type, and zone.

Source service: nodes-manager (port 8098) Endpoint: GET /api/v1/nodepools/{name}/nodeclaims

Parameters:

Name	Type	Required	Description
`pool_name`	string	Yes	NodePool name

get_scheduling_decisions

Returns the AI scheduling decision log with workload class (latency_sensitive/batch/gpu/stateful/stateless), 5-dimension score breakdown, and placement rationale.

Source service: nodes-manager (port 8098) Endpoint: GET /api/v1/decisions

Parameters:

Name	Type	Required	Description
`cluster_id`	string	No	Filter by cluster
`limit`	integer	No	Max results (default: 20)

get_spot_market

Returns spot instance pricing, savings vs on-demand, and interruption frequency by zone and instance family.

Source service: nodes-manager (port 8098) Endpoint: GET /api/v1/spot/market

Parameters:

Name	Type	Required	Description
`region`	string	No	AWS region (e.g. `us-east-1`)

Example response:

[
  {
    "instance_type": "m5.2xlarge",
    "zone": "us-east-1a",
    "current_price": 0.1212,
    "on_demand_price": 0.384,
    "savings_pct": 68.4,
    "interruption_freq": "low",
    "available": true
  }
]

get_consolidation_plan

Returns the node consolidation plan showing which nodes can be safely drained, projected cost savings, and PDB impact analysis.

Source service: nodes-manager (port 8098) Endpoint: GET /api/v1/optimize/plan

Parameters: None

get_workload_placement

Returns all workloads with their scheduled node, AI-assigned workload class, placement score (0–1 across 5 dimensions), and rationale.

Source service: nodes-manager (port 8098) Endpoint: GET /api/v1/workloads/placement

Parameters: None

Write Action Tools

These tools trigger Kubernetes mutations via action-agent-srv. All are marked ReadOnlyHint: false.

drain_node_action

Gracefully drains a Kubernetes node: cordons it then evicts all non-DaemonSet pods with a 30-second grace period.

Source service: action-agent-srv (port 8094) Endpoint: POST /api/v1/actions/drain

Parameters:

Name	Type	Required	Description
`node_name`	string	Yes	Kubernetes node name to drain

rollback_deployment

Triggers a rolling restart of a Kubernetes Deployment by patching the restartedAt annotation.

Source service: action-agent-srv (port 8094) Endpoint: POST /api/v1/actions/rollback

Parameters:

Name	Type	Required	Description
`namespace`	string	Yes	Kubernetes namespace
`deployment_name`	string	Yes	Deployment to roll back

Continuous Optimisation Tools (kubeopera-ai)

These tools call kubeopera-ai (port 8104) to access the continuous monitoring agent's data and trigger AI analysis.

get_optimization_status

Returns the latest AI-generated cluster health assessment from the kubeopera-ai monitoring agent: health score, status, severity-ranked issues with fix commands, last resource optimisation recommendations, and service uptime.

Source service: kubeopera-ai (port 8104) Endpoint: GET /api/v1/status

Parameters: None

Example response:

{
  "last_health": {
    "health_score": 68,
    "status": "warning",
    "summary": "Two pods in CrashLoopBackOff, memory pressure on worker-3",
    "issues": [
      {
        "severity": "critical",
        "component": "payments-api",
        "description": "3 pods restarting every 2 minutes",
        "recommendation": "Increase memory limits or fix OOM leak",
        "command": "kubectl describe pod -n production -l app=payments-api"
      }
    ],
    "generated_at": "2026-04-25T10:00:00Z"
  },
  "namespace": "kubeopera-prod",
  "uptime": "4h32m"
}

get_load_test_analysis

Triggers a Claude-powered load test analysis via kubeopera-ai. Reads test output and resource snapshots (before and after) from the load-generator pod, then produces a structured Markdown report with p95/p99 latency assessment, bottleneck identification, and a numbered remediation action list.

Source service: kubeopera-ai (port 8104) Endpoint: POST /api/v1/analyze/load-test

Parameters:

Name	Type	Required	Description
`results_path`	string	No	Path to results directory (uses latest if omitted)

Thresholds evaluated: p95 < 500ms, p99 < 1000ms, error rate < 1%, resource usage 60–80% of limits at peak.

Agent Launcher Tools

These tools trigger a reasoning agent run in agent-runtime (port 8111) and return a run_id. The agent then autonomously calls read-only tools to investigate, reason, and produce a structured response.

The returned run_id can be used to stream agent reasoning via the KubeOpera UI at /agents/runs/{run_id}.

run_sre_agent

Triggers a full cluster investigation by the SRE Orchestrator agent (Claude Sonnet 4.6 with interleaved thinking). Has access to all 17 read-only tools.

Parameters:

Name	Type	Required	Description
`cluster_id`	string	Yes	Cluster to investigate
`prompt`	string	Yes	Investigation goal or question

Example:

run_sre_agent(
  cluster_id: "prod-us-east",
  prompt: "Investigate the high CPU alert on payments-api and recommend immediate action"
)

Returns:

{
  "run_id": "run-abc123",
  "status": "running",
  "stream_url": "http://localhost:8111/api/v1/runs/run-abc123/stream"
}

run_security_agent

Triggers a security posture analysis by the Security Auditor agent (Claude Haiku 4.5). Uses 5 security-focused tools: get_security_posture, get_cluster_list, get_anomaly_events, get_incidents, get_agent_telemetry.

Parameters:

Name	Type	Required	Description
`cluster_id`	string	Yes	Cluster to audit
`prompt`	string	Yes	Security question or audit scope

run_cost_agent

Triggers a cost analysis by the Cost Optimizer agent (Claude Haiku 4.5). Uses 5 cost-focused tools: get_cluster_cost, get_optimization_report, get_scaling_forecasts, get_cluster_list, get_multi_cluster_overview.

Parameters:

Name	Type	Required	Description
`cluster_id`	string	Yes	Cluster to optimize
`prompt`	string	Yes	Cost analysis question

run_incident_agent

Triggers an incident triage by the Incident Responder agent (Claude Sonnet 4.6 with interleaved thinking). Can investigate, execute runbooks, and generate post-incident reports autonomously.

Parameters:

Name	Type	Required	Description
`cluster_id`	string	Yes	Cluster with the incident
`prompt`	string	Yes	Incident description or triage question

run_node_ops_agent

Triggers a node pool analysis by the NodeOps agent (Claude Sonnet 4.6 with interleaved thinking). Analyses node pool efficiency, scheduling decisions, spot market savings, and consolidation opportunities.

Parameters:

Name	Type	Required	Description
`cluster_id`	string	Yes	Cluster to analyse
`prompt`	string	Yes	Investigation goal (e.g. "Identify consolidation opportunities and spot savings")

run_load_test_agent

Triggers a load test performance analysis by the Load Test Analyst agent (Claude Sonnet 4.6). Fetches kubeopera-ai optimisation status as a baseline, triggers full load test analysis, and cross-references with live cluster metrics to produce a prioritised performance report.

Parameters:

Name	Type	Required	Description
`cluster_id`	string	Yes	Cluster to analyse
`prompt`	string	Yes	Analysis goal (e.g. "Analyse the latest load test and identify scaling bottlenecks")

App Advisor Tools

These are tenant-scoped rather than cluster-wide — each call is bound to one specific app, and the service enforces that a caller can only ever reach an app belonging to their own tenant.

get_app_profile

Fetches a discovered domain profile for one app — what kind of workload it is, its dependencies, and its observed behavior.

Source service: app-advisor-srv (port 8105) Endpoint: GET /api/v1/apps/{app_id}/profile

Parameters:

Name	Type	Required	Description
`app_id`	string	Yes	The app to profile

get_app_advice

Fetches SRE-style advice for one app — configuration and operational recommendations grounded in its discovered profile.

Source service: app-advisor-srv (port 8105) Endpoint: GET /api/v1/apps/{app_id}/advice

Parameters:

Name	Type	Required	Description
`app_id`	string	Yes	The app to advise on

run_app_advisor_agent

Triggers a full App Advisor investigation for one app, combining its profile and advice into a single reasoning pass.

Parameters:

Name	Type	Required	Description
`app_id`	string	Yes	The app to investigate
`prompt`	string	Yes	Investigation goal

Observability Gap-Fill Tools

These fill in signal types the original read-only tool set didn't cover — logs, SLOs, APM metrics, traces, and root-cause analysis each have their own dedicated backend service rather than being folded into get_cluster_health or get_pod_metrics.

get_logs

Fetches recent logs for a pod or deployment.

Source service: log-gateway (port 3100)

Parameters:

Name	Type	Required	Description
`namespace`	string	Yes	Namespace
`pod_name` \| `deployment_name`	string	One of the two	What to fetch logs for

get_slo_status

Fetches SLO compliance and error-budget burn for a service.

Source service: slo-manager (port 9090)

Parameters:

Name	Type	Required	Description
`service`	string	Yes	The service to check

get_apm_metrics

Fetches application performance metrics (latency, throughput, error rate) for a service.

Source service: apm-gateway (port 9090)

Parameters:

Name	Type	Required	Description
`service`	string	Yes	The service to query

get_traces

Fetches distributed traces for a service.

Source service: tracing-gateway (port 16686)

Parameters:

Name	Type	Required	Description
`service`	string	Yes	The service to fetch traces for

get_rca_analysis

Fetches root-cause analysis results for an incident or anomaly.

Source service: rca-engine (port 8116)

Parameters:

Name	Type	Required	Description
`incident_id`	string	Yes	The incident to analyse

GitOps Tools

These talk directly to Argo CD and a Git-hosting API rather than to another KubeOpera backend service — see Configuration for the separate env vars (ARGOCD_API_URL, ARGOCD_TOKEN, FLUX_WEBHOOK_URL, FLUX_WEBHOOK_TOKEN, GITOPS_REPO_API_URL, GITOPS_REPO_TOKEN) they read.

sync_argocd_app

Triggers an Argo CD application sync.

Parameters:

Name	Type	Required	Description
`app_name`	string	Yes	The Argo CD application to sync

get_argocd_app_status

Fetches an Argo CD application's current sync and health status.

Parameters:

Name	Type	Required	Description
`app_name`	string	Yes	The Argo CD application to check

push_gitops_patch

Commits a patch to a GitOps repository.

Parameters:

Name	Type	Required	Description
`repository`	string	Yes	Target repository
`path`	string	Yes	File path within the repository
`patch`	string	Yes	The patch content
`commit_message`	string	Yes	Commit message

trigger_flux_reconcile

Forces a Flux Kustomization to reconcile immediately, rather than waiting for its next scheduled interval.

Parameters:

Name	Type	Required	Description
`kustomization_name`	string	Yes	The Kustomization to reconcile
`namespace`	string	Yes	Its namespace

Tool Availability by Agent

The App Advisor, observability gap-fill, and GitOps tools above were added after this matrix was last updated, and their exact per-agent availability hasn't been re-confirmed since — treat their absence below as "not yet documented," not as "unavailable to every agent."

Tool	SRE	Security	Cost	Incident	NodeOps	Load Test
`get_cluster_health`	Yes	—	—	Yes	Yes	Yes
`get_cluster_cost`	Yes	—	Yes	—	—	—
`get_optimization_report`	Yes	—	Yes	—	—	—
`get_pod_metrics`	Yes	—	—	Yes	—	Yes
`get_node_metrics`	Yes	—	—	Yes	—	Yes
`get_security_posture`	Yes	Yes	—	—	—	—
`get_pipeline_status`	Yes	—	—	—	—	—
`get_cluster_list`	Yes	Yes	Yes	—	—	—
`get_anomaly_events`	Yes	Yes	—	Yes	Yes	Yes
`get_alert_rules`	Yes	—	—	—	—	—
`get_scaling_forecasts`	Yes	—	Yes	—	—	Yes
`get_incidents`	Yes	Yes	—	Yes	Yes	—
`execute_runbook`	Yes	—	—	Yes	—	—
`get_runbook_execution`	Yes	—	—	Yes	—	—
`generate_post_incident_report`	Yes	—	—	Yes	—	—
`get_multi_cluster_overview`	Yes	—	Yes	—	—	—
`get_node_pools`	Yes	—	—	—	Yes	—
`get_node_claims`	Yes	—	—	—	Yes	—
`get_scheduling_decisions`	Yes	—	—	—	Yes	—
`get_spot_market`	Yes	—	—	—	Yes	—
`get_consolidation_plan`	Yes	—	—	—	Yes	—
`get_workload_placement`	Yes	—	—	—	Yes	—
`drain_node_action`	Yes	—	—	Yes	Yes	—
`rollback_deployment`	Yes	—	—	Yes	—	—
`get_agent_telemetry`	Yes	Yes	—	Yes	—	—
`get_analysis_results`	Yes	—	—	—	—	—
`get_action_log`	Yes	—	—	Yes	—	—
`get_feedback_outcomes`	Yes	—	—	—	—	—
`get_recommendations`	Yes	—	—	Yes	—	—
`get_optimization_status`	Yes	—	—	—	—	Yes
`get_load_test_analysis`	Yes	—	—	—	—	Yes

Read-Only Data Tools​

get_cluster_health​

get_cluster_cost​

get_optimization_report​

get_pod_metrics​

get_node_metrics​

get_security_posture​

get_pipeline_status​

get_cluster_list​

get_anomaly_events​

get_scaling_forecasts​

get_incidents​

get_multi_cluster_overview​

get_agent_telemetry​

get_analysis_results​

get_action_log​

get_feedback_outcomes​

get_recommendations​

get_alert_rules​

execute_runbook​

get_runbook_execution​

generate_post_incident_report​

get_node_pools​

get_node_claims​

get_scheduling_decisions​

get_spot_market​

get_consolidation_plan​

get_workload_placement​

Write Action Tools​

drain_node_action​

rollback_deployment​

Continuous Optimisation Tools (kubeopera-ai)​

get_optimization_status​

get_load_test_analysis​

Agent Launcher Tools​

run_sre_agent​

run_security_agent​

run_cost_agent​

run_incident_agent​

run_node_ops_agent​

run_load_test_agent​

App Advisor Tools​

get_app_profile​

get_app_advice​

run_app_advisor_agent​

Observability Gap-Fill Tools​

get_logs​

get_slo_status​

get_apm_metrics​

get_traces​

get_rca_analysis​

GitOps Tools​

sync_argocd_app​

get_argocd_app_status​

push_gitops_patch​

trigger_flux_reconcile​

Tool Availability by Agent​

Read-Only Data Tools

get_cluster_health

get_cluster_cost

get_optimization_report

get_pod_metrics

get_node_metrics

get_security_posture

get_pipeline_status

get_cluster_list

get_anomaly_events

get_scaling_forecasts

get_incidents

get_multi_cluster_overview

get_agent_telemetry

get_analysis_results

get_action_log

get_feedback_outcomes

get_recommendations

get_alert_rules

execute_runbook

get_runbook_execution

generate_post_incident_report

get_node_pools

get_node_claims

get_scheduling_decisions

get_spot_market

get_consolidation_plan

get_workload_placement

Write Action Tools

drain_node_action

rollback_deployment

Continuous Optimisation Tools (kubeopera-ai)

get_optimization_status

get_load_test_analysis

Agent Launcher Tools

run_sre_agent

run_security_agent

run_cost_agent

run_incident_agent

run_node_ops_agent

run_load_test_agent

App Advisor Tools

get_app_profile

get_app_advice

run_app_advisor_agent

Observability Gap-Fill Tools

get_logs

get_slo_status

get_apm_metrics

get_traces

get_rca_analysis

GitOps Tools

sync_argocd_app

get_argocd_app_status

push_gitops_patch

trigger_flux_reconcile

Tool Availability by Agent