Check data asset health and reliability status

A Monte Carlo observability skill that checks the health of a single data asset and produces a structured health report covering freshness, active alerts, monitoring coverage, importance, and upstream dependency health. It routes all calls through the plugin's bundled MCP server using fully-qualified tool names (mcp__plugin_mc-agent-toolkit_monte-carlo-mcp__<tool>), explicitly avoiding any separately-configured monte-carlo-mcp server that might point at different credentials. Before making any tool calls it requires reading two reference files - workflows.md for exact tool calls and execution order, and parameters.md for parameter conventions. The report format is a fixed template: a header block (table name, tags, link, warehouse, status, importance score, average reads/writes per day), a metrics table (Last Activity, Alerts, Monitoring, Upstream with color-coded signals), then four always-present sections - Active Alerts (max 5 shown, overflow noted as plain text below the table), Monitors, Upstream Issues, and Recommendations - each with defined empty-state text when there's nothing to report, plus a 1-2 sentence Diagnosis. Status is determined by fixed rules: Unhealthy if any active alert exists (not-acknowledged, acknowledged, or work-in-progress), Degraded if there are no active alerts but zero active monitors on a high-importance asset (importance over 0.8 or flagged key asset), and Healthy if there are no active alerts and at least one active monitor. Each metric row is tied to a specific tool: Last Activity from get_table, Alerts from get_alerts, Monitoring from get_monitors (counting non-paused monitors), and Upstream from get_asset_lineage plus additional checks - the skill explicitly forbids inferring or fabricating any value beyond what those tools return, and only reports upstream lineage, never downstream.

Activate it when a user asks how a specific table or asset is doing, checks its status or reliability, or references it during incident triage or change planning. Do not use it for profiling or exploring table data like row counts or column stats - that's the explore-table skill; do not use it to create or suggest monitors - that's monitoring-advisor; and do not use it for active root-cause investigation of a firing alert - that's Workflow 3 of the prevent skill. This skill only reports what the defined tools return in a fixed format, so it's not the right tool when the task calls for exploration, monitor creation, or deep incident diagnosis.

Input is the name or identifier of a specific data table or asset. Output is a structured health report following the fixed template, with the health-check webapp URL always resolved dynamically via get_mc_webapp_url() rather than hardcoded, since it varies by environment.

Calls Monte Carlo's MCP tools exclusively through the plugin's bundled server (get_alerts, search, get_table, get_monitors, get_asset_lineage, get_mc_webapp_url), never a separately-configured monte-carlo-mcp server in the same session. Output formatting rules require Unicode emoji characters rather than emoji shortcodes, and bare URLs rather than markdown links, since shortcodes render as raw text in a terminal.

Data engineers and analysts who need a quick, consistently-formatted health check on a specific Monte Carlo-monitored asset - covering freshness, alerts, monitoring coverage, and upstream issues - without exploring the table's data or diagnosing a live incident.