Client Health Report

Generate a deep-dive health report for a single CRM client, displaying the overall health score, per-metric breakdowns with scoring details, risk flags, recommended actions, and recent activity timeline.

Load Skills

Read the client-health-scoring skill at ${CLAUDE_PLUGIN_ROOT}/skills/health/client-health-scoring/SKILL.md for the 5-metric scoring algorithm, RAG classification, data source integration, caching strategy, and risk flag taxonomy.

Read the sentiment-analysis skill at ${CLAUDE_PLUGIN_ROOT}/skills/health/sentiment-analysis/SKILL.md for email and meeting sentiment extraction needed by the Sentiment metric.

Parse Arguments

Extract from $ARGUMENTS:

• $1 (required) — client name to report on. Fuzzy match against CRM Companies database.

If no argument is provided:

1. Search Notion for the CRM Companies database.
2. List the first 10 active clients by name.
3. Prompt: "Which client would you like a health report for? Select from the list above or type a name."

Business Context (Optional)

Check if context files exist at _infrastructure/context/active/. If the directory contains .md files, read business-info.md, strategy.md, and current-data.md. Use this context to personalize output (e.g., prioritize known clients, use correct terminology, align with current strategy). If files don't exist, skip silently.

Step 0: Memory Context

Read the context-injection skill at _infrastructure/memory/context-injection/SKILL.md. Query for memories relevant to the current input (company, contacts, topics detected in arguments). If memories are returned, incorporate them into your working context for this execution.

Report Process

1. Identify client: Search the CRM Companies database for a company matching the provided name. Use fuzzy matching (case-insensitive, partial match). If multiple matches, list them and ask the user to clarify. If no match, report: "No client matching '[name]' found in CRM."

2. Check cache: Using the database discovery logic from the client-health-scoring skill (search "[FOS] Companies" first, then "Founder OS HQ - Companies", then "Companies", then fallback to standalone "Client Health Dashboard - Health Scores"), check the Company page for cached health scores:

   • If the Company page has a Last Scanned property < 24 hours old, use the cached scores (Health Score, Health Status, Risk Flags, per-metric scores) for the initial display. Still gather recent activity for the timeline.
   • If Last Scanned is missing or stale (> 24 hours), recompute all metrics from source data.
   • In fallback mode (standalone Health Scores DB), look for a matching record by Client Name instead.

3. Compute or retrieve metrics: For each of the 5 metrics, either use cached scores or compute fresh:

   • Last Contact (0.25 weight): Find the most recent email or calendar event with this client. Report the date and channel.
   • Response Time (0.20 weight): Calculate average response time across last 10 email threads. Report the average in hours.
   • Open Tasks (0.20 weight): Count total active tasks and overdue tasks for this client. Report the ratio.
   • Payment Status (0.20 weight): Check P11 Invoice Processor DB for this client's invoices. Report paid-on-time count vs total and any overdue invoices.
   • Sentiment (0.15 weight): Run sentiment-analysis skill. Report the composite score, email/meeting breakdowns, and any notable signals.

4. Calculate composite: Apply weights, compute overall score, classify RAG tier.

5. Evaluate risk flags: Check each risk condition from the scoring skill. List active flags.

6. Generate recommended actions: Based on risk flags and low-scoring metrics:

   • Red metrics (score < 40): Urgent action with specific recommendation
   • Yellow metrics (score 40-69): Monitoring action with suggestion
   • Green metrics (score >= 70): No action needed

7. Gather recent activity: Collect timeline data:

   • Last 5 emails involving this client (date, subject, direction, snippet)
   • Last 3 calendar events with this client (date, title, duration, attendees)

8. Update cache: If scores were recomputed, update the Company page with the new health score properties (Health Score, Health Status, Risk Flags, Last Scanned, Sources Used, Notes). In fallback mode, write/update the record in the standalone Health Scores DB.

Graceful Degradation

• Notion unavailable: Cannot identify client. Report error and stop.
• Gmail unavailable: Set Last Contact and Response Time to defaults. Note in report.
• Calendar unavailable: Skip meeting data. Use email-only for timeline and sentiment.
• P11 unavailable: Set Payment to neutral (75). Note in report.
• P20 dossier available: Supplement the report with dossier data if fresh.

Output Format

Display the complete health report in chat:

## Health Report: [Client Name]

**Overall Score**: [score]/100 — [Green 🟢 | Yellow 🟡 | Red 🔴]
**Last Scanned**: [timestamp] | **Data Sources**: [list of sources used]

---

### Metric Breakdown

| Metric | Score | Weight | Details |
|--------|-------|--------|---------|
| Last Contact | [score]/100 | 0.25 | Last contact: [date] via [email/meeting] ([N] days ago) |
| Response Time | [score]/100 | 0.20 | Avg response: [N] hours across [N] threads |
| Open Tasks | [score]/100 | 0.20 | [N] active tasks, [N] overdue |
| Payment Status | [score]/100 | 0.20 | [N]/[N] invoices paid on time, [N] overdue |
| Sentiment | [score]/100 | 0.15 | [Positive/Neutral/Negative] — Email: [score], Meeting: [score] |

### Risk Flags

[If no flags]: No risk flags detected.
[If flags exist]:
- 🔴 **[Flag Name]**: [Description and trigger condition]
- 🟡 **[Flag Name]**: [Description and trigger condition]

### Recommended Actions

[Based on risk flags and low-scoring metrics]:
1. **[Action]** — [Specific recommendation based on the data]
2. **[Action]** — [Specific recommendation based on the data]

### Recent Activity

**Emails** (last 5):
- [date] — [↗ Sent | ↙ Received] "[subject]" — [first 80 chars of body]

**Meetings** (last 3):
- [date] — "[title]" ([duration]) — Attendees: [names]

---

Report generated from [N] data sources. Use `/founder-os:client:health-scan --client="[name]" --refresh` to force a full rescan.

Final: Memory Update

Read the pattern-detection skill at _infrastructure/memory/pattern-detection/SKILL.md. Log this execution as an observation with: plugin name, primary action performed, key entities (companies, contacts), and output summary. Check for emerging patterns per the detection rules. If a memory reaches the adaptation threshold, append the notification to the output.

Usage Examples

/founder-os:client:health-report Acme Corp
/founder-os:client:health-report "Smith & Associates"
/founder-os:client:health-report acme