regenerate-fixtures

name	regenerate-fixtures
description	Regenerate all test fixtures and snapshot data from current pricing configuration. Destructive — overwrites existing test data.
disable-model-invocation	true
allowed-tools	Read, Write, Edit, Bash, Glob, Grep

Regenerate Test Fixtures

This skill regenerates all test fixture data from the current pricing configuration. It overwrites existing test files, so it should only be run deliberately.

Invocation: /regenerate-fixtures — this skill will never auto-trigger.

Steps

1. Read Current Pricing Config

Load all pricing data from src/data/:

• pricing.ts — commitment tiers, PAYG rates
• logSources.ts — log source definitions and ingestion ranges
• defenderMapping.ts — M365 licence to Defender coverage mapping

2. Generate Calculation Fixtures

Create test fixture files in src/__fixtures__/ with pre-computed expected values:

payg-fixtures.json:

[
  { "gbPerDay": 0, "monthlyUsd": 0, "monthlyGbp": 0 },
  { "gbPerDay": 1, "monthlyUsd": 158.29, "monthlyGbp": 125.05 },
  { "gbPerDay": 10, "monthlyUsd": 1582.88, "monthlyGbp": 1250.48 },
  { "gbPerDay": 50, "monthlyUsd": 7914.40, "monthlyGbp": 6252.38 },
  { "gbPerDay": 100, "monthlyUsd": 15828.80, "monthlyGbp": 12504.75 },
  { "gbPerDay": 500, "monthlyUsd": 79144.00, "monthlyGbp": 62523.76 }
]

tier-fixtures.json: For each commitment tier, generate:

• Monthly cost at exactly the tier level
• Monthly cost at 50% of tier level (underage)
• Monthly cost at 150% of tier level (overage)
• Effective rate per GB
• Breakeven GB/day vs PAYG
• Savings percentage vs PAYG at tier level

recommendation-fixtures.json: For ingestion values at 5, 10, 25, 40, 60, 70, 100, 150, 250, 400, 800, 1500, 3000 GB/day:

• Best tier name
• Monthly cost of best tier
• Monthly cost of PAYG for comparison
• Savings percentage

defender-overlap-fixtures.json: For each licence type (E3, E5, E5 Security), with a standard set of enabled sources:

• Total ingestion before overlap
• Total ingestion after overlap
• GB/day reduction
• Monthly USD savings

3. Generate Snapshot Data

Create src/__fixtures__/snapshot-inputs.json with 5 representative customer profiles:

1. Small E3 customer — 200 users, basic sources, no Defender overlap
2. Mid-market E5 customer — 1,000 users, full Microsoft stack, Defender overlap applied
3. Large E5 customer — 3,000 users, all sources including network, Defender overlap
4. E3 + heavy third-party — 500 users, third-party firewall dominant
5. Minimal deployment — 100 users, only Entra ID and Activity Logs

For each profile, compute and store the expected:

• Total GB/day, billable GB/day, free GB/day
• PAYG monthly cost
• Best tier and its monthly cost
• Savings vs PAYG

4. Update Test Files

If test files reference hardcoded expected values, update them to match the new fixtures.

5. Verify

Run npm test after regeneration to confirm all tests still pass with the updated fixtures.

Output

Report:

• Number of fixture files created/updated
• Any tests that failed after regeneration
• Summary of pricing values used to generate fixtures