Use this skill whenever the user needs VMware Aria Operations data — performance metrics, alerts, capacity planning, anomaly detection, and automated reports. Directly handles: query resource metrics, list/acknowledge/cancel alerts, manage alert definitions, check capacity and time-remaining forecasts, detect anomalies, generate and manage reports. Always use this skill for "check vSphere capacity", "what Aria Operations alerts are active", "show VMware anomalies", "generate an Aria report", "rightsizing recommendations", or any Aria/vRealize Operations task. Combined with LLM, Aria data powers natural language reports: "give me a capacity report" → Aria collects data → LLM formats the report. Do NOT use for real-time vCenter alarms/events (use vmware-monitor), VM operations (use vmware-aiops), or NSX networking (use vmware-nsx). For load balancing/AVI/AKO use vmware-avi.
Install with Codex or Claude Copy this prompt, paste it into Codex, Claude, or another assistant, and let it review the skill page and install it for you.
A direct command skips the review prompt. Inspect the source before running it.
Use this skill whenever the user needs VMware Aria Operations data — performance metrics, alerts, capacity planning, anomaly detection, and automated reports. Directly handles: query resource metrics, list/acknowledge/cancel alerts, manage alert definitions, check capacity and time-remaining forecasts, detect anomalies, generate and manage reports. Always use this skill for "check vSphere capacity", "what Aria Operations alerts are active", "show VMware anomalies", "generate an Aria report", "rightsizing recommendations", or any Aria/vRealize Operations task. Combined with LLM, Aria data powers natural language reports: "give me a capacity report" → Aria collects data → LLM formats the report. Do NOT use for real-time vCenter alarms/events (use vmware-monitor), VM operations (use vmware-aiops), or NSX networking (use vmware-nsx). For load balancing/AVI/AKO use vmware-avi.
vmware-policy auto-installed as Python dependency (provides @vmware_tool decorator and audit logging). All write operations audited to ~/.vmware/audit.db. Credentials: Each Aria Operations target requires a per-target password env var in ~/.vmware-aria/.env following the pattern VMWARE_ARIA_<TARGET_NAME_UPPER>_PASSWORD. Passwords are never logged or echoed. Read-heavy: 21 of 28 tools are read-only. Write operations limited to alert acknowledge/cancel, alert definition management, and report management. No webhooks, no outbound network calls, no guest operations. Local only: stdio MCP + Aria Operations REST API (HTTPS 443). Transitive dependencies: Only vmware-policy (audit/policy). No post-install scripts or background services.
VMware Aria Operations
Disclaimer: This is a community-maintained open-source project and is not affiliated with, endorsed by, or sponsored by VMware, Inc. or Broadcom Inc. "VMware" and "Aria" are trademarks of Broadcom. Source code is publicly auditable at github.com/zw008/VMware-Aria under the MIT license.
Diagnostic investigations: Before running any "why is X slow / failing / down" workflow, follow references/investigation-protocol.md. It enforces the four root-cause completeness criteria (falsifiability / sufficiency / necessity / mechanism) and the up-to-three-rounds deepening loop. Stopping at a partial conclusion is an anti-pattern — always self-check against the criteria before outputting a report.
Daily VM Health Check (Proactive Ops)
Judgment: don't chase the highest CPU consumer — chase the highest contention consumer. A VM at 90% CPU on a quiet host is healthy; a VM at 30% CPU but 15% Ready is starving. Key metrics: CPU Ready, Memory Balloon, Disk Latency.
Find top CPU consumers → vmware-aria resource top --metric cpu|usage_average --top 20 (this is the starting set, not the answer)
Check CPU Ready on hot VMs → vmware-aria resource metrics <vm-id> --metrics cpu.ready.summation --hours 24
List active CRITICAL/IMMEDIATE alerts → vmware-aria alert list --criticality CRITICAL
Check anomaly counts → vmware-aria anomaly list
Cross-validate against the investigation protocol before reporting any "root cause" — high consumption is rarely the root, usually a downstream symptom
Investigate High CPU Alert
List active CRITICAL alerts → vmware-aria alert list --criticality CRITICAL
Get alert details + symptoms → vmware-aria alert get <alert-id>
Find top CPU consumers → vmware-aria resource top --metric cpu|usage_average
Fetch 24h CPU metrics for the hot VM → vmware-aria resource metrics <vm-id> --metrics cpu|usage_average --hours 24
Create alert definition referencing symptoms → MCP create_alert_definition with name, resource_kind, symptom_definition_ids, criticality (any one symptom firing triggers the alert)
Verify it appears in definitions → vmware-aria alert definitions --name "Gold VM CPU" (criticality shown is the max severity across the definition's states)
Enable or disable later → MCP set_alert_definition_state
Trigger report generation → vmware-aria report generate <definition-id> --resources <resource-id> (the Report API requires at least one resource UUID)
Poll until completed → vmware-aria report get <report-id> (repeat until status == COMPLETED)
Download via the returned download_url (PDF) or csv_url
Clean up → vmware-aria report delete <report-id>
Multi-Target Operations
All commands accept --target <name> to operate against a specific Aria Ops instance:
vmware-aria alert list --target prod
vmware-aria resource top --target lab
Usage Mode
Scenario
Recommended
Why
Local/small models (Ollama, Qwen)
CLI
~2K tokens vs ~8K for MCP
Cloud models (Claude, GPT-4o)
Either
MCP gives structured JSON I/O
Automated pipelines
MCP
Type-safe parameters, structured output
Running vmware-aria with a local or small model? See references/agent-guardrails.md for tool-calling guardrails (alert-to-resource correlation and Aria data fidelity).
MCP Tools (28 — 21 read, 7 write)
All MCP tools accept an optional target parameter to select which Aria Operations instance to connect to.
Category
Tool
Type
Description
Resource
list_resources
Read
List VMs, hosts, clusters by resource kind
get_resource
Read
Get resource details with health, risk, efficiency badges
get_resource_metrics
Read
Fetch time-series metric stats for any resource
get_resource_health
Read
Get health badge score (0–100)
get_top_consumers
Read
Rank resources by CPU, memory, disk, or network usage
Alerts
list_alerts
Read
List active alerts with criticality and resource ID (resolve names via get_resource)
get_alert
Read
Get alert details with contributing symptoms (recommendations live on the alert definition)
investigate_alert
Read
Resolve an alert to its confirmed affected resource in one call — returns both UUIDs explicitly labelled plus the vmware-monitor handoff
acknowledge_alert
Write
Mark an alert as acknowledged (does not close it)
cancel_alert
Write
Cancel (dismiss) an active alert
list_alert_definitions
Read
List alert templates configured in Aria Ops
Alert Defs
list_symptom_definitions
Read
List symptom definitions — use IDs when creating alert defs
create_alert_definition
Write
Create new alert definition from symptom definition IDs
set_alert_definition_state
Write
Enable or disable an alert definition
delete_alert_definition
Write
Delete an alert definition permanently
Capacity
get_capacity_overview
Read
Group-level remaining % + per-dimension headroom and days-until-full
Read/write split: 21 read-only, 7 write. All write operations are audit-logged to ~/.vmware/audit.db (via vmware-policy).
List results are envelopes — read truncated before you summarise
Every list-returning tool above (list_resources, get_top_consumers, list_alerts, list_alert_definitions, list_symptom_definitions, list_rightsizing_recommendations, list_report_definitions, list_reports, list_anomalies, list_collector_groups) returns an object, not a bare array:
{"items":[ ... ],"returned":50,"limit":50,"total":213,"truncated":true,"hint":"Showing 50 of 213. Raise limit or narrow the query with a filter to see the rest."}
Rules:
Rows live under items. An empty items with returned: 0 means the query genuinely matched nothing — report that, do not report a tool failure.
truncated: true means more rows exist. Never describe such a result as the complete set; either say it is a partial view or re-query with a higher limit or a narrower filter, as hint instructs.
truncated: false means the answer is complete — safe to summarise as the whole picture.
total: null means the API reported no collection size, so a page filled exactly to the limit is flagged truncated conservatively. It may in fact be complete; a follow-up query with a larger limit settles it.
list_anomalies also carries scanned — how many VMs were examined. It returns only VMs with a non-zero anomaly count, so a short list is not evidence that the environment is clean; check truncated.
CLI Quick Reference
# Resources
vmware-aria resource list [--kind VirtualMachine|HostSystem|ClusterComputeResource] [--name <filter>]
vmware-aria resource get <resource-id>
vmware-aria resource metrics <resource-id> --metrics cpu|usage_average,mem|usage_average --hours 4
vmware-aria resource metrics <vm-id> --metrics cpu.ready.summation,mem.balloon.average --hours 24
vmware-aria resource health <resource-id>
vmware-aria resource top --metric cpu|usage_average --kind VirtualMachine --top 10
# Alerts
vmware-aria alert list [--criticality CRITICAL|IMMEDIATE|WARNING|INFORMATION]
vmware-aria alert get <alert-id>
vmware-aria alert acknowledge <alert-id>
vmware-aria alert cancel <alert-id>
vmware-aria alert definitions [--name <filter>]
# Alert Definitions: creation/enable/disable/delete and symptom-definition# lookup are MCP-only tools (list_symptom_definitions, create_alert_definition,# set_alert_definition_state, delete_alert_definition) — no CLI subcommands.# Capacity
vmware-aria capacity overview <cluster-id>
vmware-aria capacity remaining <resource-id>
vmware-aria capacity time-remaining <resource-id>
vmware-aria capacity rightsizing [--resource-id <vm-id>]
# Reports (async: generate → poll get → download → delete)
vmware-aria report definitions [--name <filter>]
vmware-aria report generate <definition-id> --resources <id1,id2> # at least one resource UUID required
vmware-aria report list [--definition-id <id>]
vmware-aria report get <report-id> # poll until status == COMPLETED; shows download_url
vmware-aria report delete <report-id>
# Anomaly
vmware-aria anomaly list [--resource-id <id>]
vmware-aria anomaly risk <resource-id>
# Health
vmware-aria health status
vmware-aria health collectors
# Diagnostics
vmware-aria doctor [--skip-auth]
Key Metric Names (for resource metrics command)
Metric
API Key
What It Means
CPU Ready %
cpu.ready.summation
vCPU waiting for physical core; >5% = warning
CPU Used
cpu.used.summation
Actual CPU execution time
CPU Demand
cpu.demand.average
Total MHz requested by VM
Memory Active
mem.active.average
Actively used by guest OS (sizing)
Memory Consumed
mem.consumed.average
Footprint on host (capacity)
Memory Balloon
mem.balloon.average
>0 = ESXi reclaiming memory
Memory Swap
mem.swapped.average
>0 = severe pressure
Disk Read Latency
disk.read.average
Read I/O latency ms
Disk Write Latency
disk.write.average
Write I/O latency ms
Net Received
net.received.average
Inbound network KB/s
Net Transmitted
net.transmitted.average
Outbound network KB/s
Full CLI reference with all options and output formats: see references/cli-reference.md
Troubleshooting
"Token not found" error after setup
The token acquisition request failed. Verify:
Aria Ops is reachable: vmware-aria doctor
The auth_source in config matches your environment (LOCAL, LDAP, AD)
The password env var follows the naming convention: VMWARE_ARIA_<TARGET>_PASSWORD
Resources appear missing from list_resources
The collector agent may be offline. Check list_collector_groups for any collectors in a DOWN state. Restart the affected collector from the Aria Ops UI under Administration > Collector Groups.
Metrics return empty data
The resource may not have metric collection configured, or the requested metric key is incorrect. Verify metric keys against the resource's available metrics in the Aria Ops UI (Metrics tab on the resource detail page).
"Password not found" error
Variable names follow the pattern VMWARE_ARIA_<TARGET_NAME_UPPER>_PASSWORD where hyphens become underscores. Example: target prod needs VMWARE_ARIA_PROD_PASSWORD. Check your ~/.vmware-aria/.env file.
invalid peer certificate: UnknownIssuer when running uvx (corporate TLS proxy)
uvx re-resolves dependencies from PyPI on every launch. Behind a corporate TLS-intercepting proxy whose CA is not in uv's bundled cert store, the handshake fails. Use the v1.5.15+ recommended single-command form vmware-aria mcp (after uv tool install vmware-aria — no network on launch), or set UV_NATIVE_TLS=true to make uv use the system cert store.
Safety
Read-heavy: 21 of 28 tools are read-only
Audit logging: Write operations logged to ~/.vmware/audit.db (SQLite WAL, via vmware-policy) with timestamp, user, target, operation, and result
Token expiry handling: vRealizeOpsToken re-acquired automatically 60 seconds before expiry (6-hour sliding validity, extended on each call)
Prompt injection defense: API text values sanitized via _sanitize() — strips control characters, truncates to 500 chars
Credential safety: Passwords loaded only from environment variables (.env file), never from config.yaml
Input validation: resource_id and alert_id validated before API calls; criticality values validated against known enum
Setup
uv tool install vmware-aria
mkdir -p ~/.vmware-aria
cp config.example.yaml ~/.vmware-aria/config.yaml
# Edit config.yaml with your Aria Operations host details# Add to ~/.vmware-aria/.env (create if missing, chmod 600):# VMWARE_ARIA_PROD_PASSWORD=<your-password>chmod 600 ~/.vmware-aria/.env
vmware-aria doctor
All tools are automatically audited via vmware-policy. Audit logs: vmware-audit log --last 20
Full setup guide with multi-target config, MCP server setup, and Docker: see references/setup-guide.md
Architecture
User (natural language)
|
AI Agent (Claude Code / Goose / Cursor)
| reads SKILL.md
vmware-aria CLI or MCP server (stdio transport)
| Aria Operations Suite API (REST/JSON over HTTPS)
| POST /suite-api/api/auth/token/acquire → vRealizeOpsToken
Aria Operations Manager
|
VMs / Hosts / Clusters / Datastores / Alerts / Capacity
The MCP server uses stdio transport (local only, no network listener). Connections to Aria Ops use HTTPS on port 443 with vRealizeOpsToken authentication (6-hour sliding token validity, auto-refreshed).
Audit & Safety
All operations are automatically audited via vmware-policy (@vmware_tool decorator):
Every tool call logged to ~/.vmware/audit.db (SQLite, framework-agnostic)
