home / skills / phrazzld / claude-config / dashboard

dashboard skill

/skills/dashboard

This skill aggregates traffic, revenue, and errors from all MistyStep products into a single weekly dashboard, highlighting traction signals for quick action.

npx playbooks add skill phrazzld/claude-config --skill dashboard

Review the files below or copy the command above to add this skill to your agents.

Files (5)
SKILL.md
5.9 KB
---
name: dashboard
description: |
  Unified portfolio dashboard. Traffic, revenue, errors - all products at a glance.
  Weekly traction check in 2 minutes.

  Auto-invoke when: user asks about metrics, traction, analytics, revenue across products.
argument-hint: "[--html to export to file]"
effort: high
---

# /dashboard

See everything. Find traction signals.

## What This Does

Aggregates metrics from all MistyStep products into a single view:
- **Traffic**: PostHog analytics per project (with MCP)
- **Revenue**: Stripe (subscriptions, one-time payments)
- **Errors**: Sentry error counts
- **Changes**: Week-over-week deltas with anomaly flags

## Output Format

```
┌──────────────────────────────────────────────────────────────────┐
│ MISTYSTEP PORTFOLIO - Week of Jan 20                            │
├──────────────────────────────────────────────────────────────────┤
│ Product          │ Visits │ Δ    │ Revenue │ Errors │ Status    │
├──────────────────────────────────────────────────────────────────┤
│ Volume           │   892  │ ↑147%│  $45.00 │   0    │ ⚠️ SIGNAL │
│ Heartbeat        │   234  │ ↑ 12%│   $0.00 │   2    │ 🟢        │
│ Scry             │    45  │ ↓  5%│   $0.00 │   0    │ 🟢        │
│ Crondle          │   156  │ →  0%│   $0.00 │   0    │ 🟢        │
│ ...              │        │      │         │        │           │
├──────────────────────────────────────────────────────────────────┤
│ TOTAL            │  2,847 │ ↑ 34%│ $127.00 │   3    │           │
└──────────────────────────────────────────────────────────────────┘

⚠️ Volume showing unusual growth (+147%)
   Investigate: What's driving traffic?
```

## Process

### 1. Load Product Registry

Read `products.yaml` for the list of products and their API identifiers:
- PostHog project IDs
- Stripe account/product IDs
- Sentry project slugs
- Domain names

### 2. Fetch Metrics

For each product, fetch:

**Traffic (PostHog API / MCP)**
```bash
# Via PostHog MCP (preferred - Claude can query directly)
# Or via CLI:
curl -s "https://app.posthog.com/api/projects/${PROJECT_ID}/insights/trend/" \
  -H "Authorization: Bearer $POSTHOG_API_KEY" \
  -d '{"events": [{"id": "$pageview"}], "date_from": "-7d"}'

# For specific breakdown by referrer:
curl -s "https://app.posthog.com/api/projects/${PROJECT_ID}/insights/trend/" \
  -H "Authorization: Bearer $POSTHOG_API_KEY" \
  -d '{"events": [{"id": "$pageview"}], "breakdown": "$referrer", "date_from": "-7d"}'
```

**Revenue (Stripe API)**
```bash
# Requires STRIPE_SECRET_KEY env var
stripe balance_transactions list --created[gte]=$LAST_WEEK --limit=100
# Or for specific product revenue:
stripe invoices list --created[gte]=$LAST_WEEK --status=paid
```

**Errors (Sentry API)**
```bash
# Requires SENTRY_AUTH_TOKEN env var
curl -s "https://sentry.io/api/0/projects/${ORG}/${PROJECT}/stats/" \
  -H "Authorization: Bearer $SENTRY_AUTH_TOKEN"
```

### 3. Calculate Deltas

Compare this week vs last week:
- ↑ = increase > 5%
- ↓ = decrease > 5%
- → = stable (within 5%)
- ⚠️ SIGNAL = change > 50% (investigate!)

### 4. Display Results

Output formatted table to terminal.

**Optional**: `--html` flag exports to `~/dashboard.html` for browser viewing.

## Product Registry

Edit `~/.claude/skills/dashboard/products.yaml` to configure products:

```yaml
products:
  - name: Volume
    domain: volume.app
    posthog_project_id: 12345  # Changed from vercel_project_id
    stripe_product_id: prod_xxx
    sentry_project: volume

  - name: Heartbeat
    domain: heartbeat.app
    posthog_project_id: 12345
    stripe_product_id: prod_yyy
    sentry_project: heartbeat

  # ... more products
```

## Required Environment Variables

```bash
# PostHog - for traffic analytics (preferred)
POSTHOG_API_KEY=phx_xxx
POSTHOG_PROJECT_ID=12345

# Stripe - for revenue
STRIPE_SECRET_KEY=sk_live_xxx

# Sentry - for errors (optional)
SENTRY_AUTH_TOKEN=xxx
SENTRY_ORG=mistystep
```

**Note:** Vercel Analytics is NOT used because it lacks CLI/API access. PostHog provides equivalent traffic data with MCP integration.

## Usage

```bash
# Weekly check (terminal output)
/dashboard

# Export to HTML for browser
/dashboard --html

# Check specific product only
/dashboard volume
```

## When to Use

- **Weekly ritual**: Every Monday, run `/dashboard` to spot traction
- **After launch**: Check if announcement drove traffic
- **Investigating issues**: See if errors spiked

## Traction Signals

The dashboard flags products with >50% change as potential traction signals.

When you see ⚠️ SIGNAL:
1. Investigate the source (referrer data)
2. Is it sustainable or a one-time spike?
3. If sustainable, consider `/double-down`

## Fallback: Browser Automation

If APIs aren't configured, fall back to browser automation:
1. Open PostHog dashboard in Chrome
2. Navigate to each project's analytics
3. Screenshot or extract numbers
4. Repeat for Stripe and Sentry dashboards

This is slower but works without API setup.

## Related Skills

- `/observability` - Full monitoring setup (Sentry, PostHog, etc.)
- `/marketing-status` - Marketing-focused metrics with PostHog MCP
- `/stripe` - Stripe integration audit
- `/double-down` - What to do when traction appears (future skill)

Overview

This skill provides a unified portfolio dashboard that surfaces traffic, revenue, and error trends across all products. It aggregates PostHog traffic, Stripe revenue, and Sentry errors into a single table with week-over-week deltas and anomaly flags. Use it for a quick weekly traction check or to investigate sudden changes across products.

How this skill works

The dashboard loads a product registry (products.yaml) with API identifiers, then queries PostHog for pageviews, Stripe for paid invoices/balance transactions, and Sentry for error counts. It calculates week-over-week deltas, applies thresholds for increases/decreases/anomalies, and outputs a compact table to the terminal or an optional HTML export. Large changes (>50%) are flagged as SIGNAL for follow-up.

When to use it

  • Weekly portfolio check (e.g., Monday ritual)
  • After a launch, campaign, or announcement to validate impact
  • When troubleshooting spikes in errors or sudden revenue changes
  • To compare product traction side-by-side for prioritization
  • Before planning resource or marketing allocations

Best practices

  • Maintain a current products.yaml with PostHog, Stripe, and Sentry identifiers
  • Keep API keys and environment variables secure and scoped (POSTHOG_API_KEY, STRIPE_SECRET_KEY, SENTRY_AUTH_TOKEN)
  • Run the dashboard both aggregate and per-product when a signal appears to narrow the source
  • Use the referrer breakdown from PostHog when investigating traffic spikes
  • Export HTML for sharing with stakeholders and attach referrer/context notes

Example use cases

  • Weekly 2-minute portfolio check to spot which products gained or lost traction
  • Post-announcement review to confirm whether a campaign drove sustained traffic or a one-off spike
  • Error triage: identify which product(s) spiked in Sentry and correlate with traffic or recent deploys
  • Revenue reconciliation: compare Stripe paid invoices week-over-week across products
  • Discovery: detect a product with >>50% growth and trigger a follow-up growth experiment

FAQ

What environment variables are required?

Provide POSTHOG_API_KEY (and project IDs in products.yaml), STRIPE_SECRET_KEY for revenue, and optionally SENTRY_AUTH_TOKEN and SENTRY_ORG for errors.

What if APIs are not configured?

The skill can fall back to browser automation: open PostHog, Stripe, and Sentry dashboards and extract numbers manually, then render the table.