adm_health_score

adm_health_scoreTier 2Writes to health_scores

Composite 0–100 health score from usage rate, sentiment, engagement breadth, and support volume. Pure computation — no LLM calls.

How to Invoke

Cursor: /adm-health [account]— No MCP required

CLI: adm health <name> [--interactive] [--licensed <n>] [--weekly-active <n>]

MCP: When MCP is enabled, ask in natural language (e.g. “Research Adobe”) and the AI invokes adm_health_score

Relationships

Feeds into: adm_risk_detector, adm_qbr_generator, adm_team_dashboard

When to Use

Use this tool after receiving usage data from your account champion — typically weekly or after a major event (QBR, new team rollout, exec review). The output persists to MongoDB and feeds adm_risk_detector, adm_qbr_generator, and adm_team_dashboard.

Do not use for accounts where you have no usage data. The output becomes meaningless without real figures.

Inputs

Parameter	Type	Description
accountId*	string	Account identifier (company name or MongoDB _id) "Adobe"
usageData*	object	Snapshot of current usage metrics { totalLicensed: 1200, weeklyActive: 980, dailyActive: 720, avgRequestsPerUserPerDay: 15 }
sentimentData	object	NPS and support metrics if available { npsScore: 45, supportTicketsCritical: 0 }
engagementData	object	Champion activity, exec sponsor contact { championActivePct: 0.8 }

UsageData Shape

{
  totalLicensed: number;   // Total licensed seats
  weeklyActive: number;   // Weekly active users
  dailyActive: number;    // Daily active users
  avgRequestsPerUserPerDay: number;
  featureDistribution?: Record<string, number>;  // Optional
}

Output

Branded markdown report with composite score, category breakdown, risk level, flags, and recommended actions.markdown

### ▸ Cursor ADM Toolkit · Health Score

**Adobe** · 🟢 LOW · 81/100

**Category breakdown**
| Category | Score | Weight |
|:---|:---|:---|
| Usage | 82 | 30% |
| Depth | 75 | 25% |
| Sentiment | 73 | 15% |
| Engagement | 90 | 15% |
| Velocity | 50 | 15% |

**Recommended actions**
• [this_quarter] Feature adoption accelerator for top teams

How the Score Is Calculated

The composite score is a weighted sum of five dimensions:

Dimension	Weight	Source
Usage rate (WAU / total)	30%	usageData
Feature depth	25%	usageData.featureDistribution
NPS / sentiment	15%	sentimentData (defaults to 50 if absent)
Engagement (champions, sponsor)	15%	engagementData
Velocity (placeholder)	15%	Default 50 until historical data

Full formula: mcp-server/lib/constants.js and handler logic in handlers.js.

CLI

adm health <name> [--interactive] [--licensed <n>] [--weekly-active <n>] [--nps <n>] [--tickets <n>] ...

Use --interactive to prompt for all inputs.

Persistence

🗄health_scoreswrites

Each call appends a timestamped document. Previous scores are retained — adm_risk_detector uses the history to detect trends.

Related Tools

• adm_risk_detector — reads health history to flag churn signals
• adm_qbr_generator — uses health scores for QBR health section
• adm_team_dashboard — aggregates scores across your portfolio