shipkit-metrics

shipkit-metrics - Metric Capture Utility

Purpose: Let users record current metric values into .shipkit/metrics/latest.json so that /shipkit-product-goals --evaluate and /shipkit-engineering-goals --evaluate can compare actuals against targets.

This is a manual capture tool — metrics come from real-world sources (analytics, APMs, test runs, user feedback) that no skill can auto-generate.

---

When to Invoke

User triggers:

• "Log metrics", "Update metrics", "Record measurements"
• "Our response time is now 200ms"
• "Conversion rate hit 12%"
• "Test coverage is at 85%"

Workflow position:

• After goals are defined (/shipkit-product-goals, /shipkit-engineering-goals)
• Before goal evaluation (--evaluate mode on goals skills)

---

Prerequisites

Recommended (to know which metrics to capture):

• .shipkit/goals/product.json — P-* criteria with thresholds
• .shipkit/goals/engineering.json — E-* criteria with thresholds

If missing: Accept any metric the user provides, using freeform IDs.

---

Arguments

• No args or metric text → Interactive mode (show goals, ask for values)
• --list → Show current metrics and which criteria are still unmeasured
• Metric ID + value (e.g., P-001 82%) → Quick capture mode

---

Process

Mode 1: Interactive (default)

1. Read .shipkit/goals/product.json and .shipkit/goals/engineering.json if they exist
2. Extract all criteria IDs and their metrics/thresholds
3. Read .shipkit/metrics/latest.json if exists (to show what's already captured)
4. Present unmeasured criteria:

Metrics Capture — {N} criteria defined, {M} measured

Unmeasured:
  P-001: Task completion rate (target: > 80%)
  P-003: Page load satisfaction (target: < 3s)
  E-002: API response time (target: < 500ms)

Already measured:
  E-001: Test coverage = 85% (target: > 80%) — exceeded
  P-002: Error rate = 2.1% (target: < 5%) — at-threshold

Enter a criterion ID and its current value (e.g., "P-001 82%"), or "done" to finish: