frontend/src/lib/mockData.ts

// ──────────────────────────────────────────────────────────────────────────
// mockData.ts
//
// Frontend-only fixtures for the Extreme Heat Insurance dashboard. The
// Vercel serverless endpoints under `api/*` hit a Neon Postgres database
// in production, but `npm run dev` doesn't run those handlers, so every
// page would otherwise be stuck on loading spinners.
//
// Instead of standing up a mock server, `src/lib/api.ts` reads directly
// from the exports below. Shapes are matched field-for-field to the real
// API response contracts so the UI needs zero adaptation.
//
// Context: Dar es Salaam, Tanzania. 15 monitored settlements. Parametric
// WBGT trigger at the city-wide P97 threshold (38.8 °C). M-Pesa / Tigo
// Pesa / Airtel Money disbursements.
// ──────────────────────────────────────────────────────────────────────────

import type {
  ZonesResponse,
  Zone,
  IndicesResponse,
  IndexData,
  DailyHeat,
  TriggersResponse,
  Trigger,
  BasisRiskResponse,
  BasisRisk,
  NotificationsResponse,
  Notification,
  PipelineRunsResponse,
  PipelineRun,
  PipelineStep,
  PipelineStats,
  EnrolledResponse,
  CoverageResponse,
  CoverageZone,
  CalibrateResponse,
  CalibrateZoneResult,
  CalibrateParams,
} from './api'
import type { DisbursementsResponse, DisbursementRecord } from '../types'

// ── Time helpers ──────────────────────────────────────────────────────────
// Deterministic "now" so repeated renders don't drift. The date is locked
// to mid-April 2026, which is late rainy-season in Dar — hot, humid, and
// inside the parametric product's peak coverage window.

const NOW = new Date('2026-04-13T08:42:00Z')

function isoDaysAgo(days: number, hour = 14, minute = 18): string {
  const d = new Date(NOW.getTime() - days * 86_400_000)
  d.setUTCHours(hour, minute, 0, 0)
  return d.toISOString()
}

function dateOnly(days: number): string {
  const d = new Date(NOW.getTime() - days * 86_400_000)
  return d.toISOString().slice(0, 10)
}

// Tiny seeded PRNG so daily-history values look natural but stay stable
// between hot reloads. Mulberry32.
function rng(seed: number) {
  let t = seed
  return () => {
    t = (t + 0x6d2b79f5) | 0
    let r = Math.imul(t ^ (t >>> 15), 1 | t)
    r = (r + Math.imul(r ^ (r >>> 7), 61 | r)) ^ r
    return ((r ^ (r >>> 14)) >>> 0) / 4294967296
  }
}

// ── Zone seed data ────────────────────────────────────────────────────────
// Real Dar es Salaam neighborhoods. Centroids are hand-checked against
// OpenStreetMap; they don't need to be pixel-precise — they exist so the
// frontend has values to render. Worker populations are drawn from ILO
// informal-sector ratios and the Tanzania NBS 2024 census.

interface ZoneSeed {
  zone_id: string
  name: string
  settlement_type: 'informal' | 'mixed' | 'formal'
  heat_vulnerability: 'low' | 'medium' | 'high' | 'severe'
  latitude: number
  longitude: number
  elevation_m: number
  worker_population_est: number
  outdoor_exposure_pct: number
  // Current conditions
  corrected_temp_c: number
  current_wbgt_c: number
  grid_temp_c: number
  uhi_delta_c: number
  // Forecast
  trigger_probability_7d: number
  model_tier: 'ensemble' | 'full_model' | 'persistence' | 'climatology'
  consecutive_hot_days: number
  // Historical
  annual_trigger_days: number
}

const ZONE_SEEDS: ZoneSeed[] = [
  // ── Informal settlements — hottest, highest worker counts
  {
    zone_id: 'DAR-MAB', name: 'Mabibo', settlement_type: 'informal', heat_vulnerability: 'severe',
    latitude: -6.798, longitude: 39.232, elevation_m: 28, worker_population_est: 11_950,
    outdoor_exposure_pct: 0.78, corrected_temp_c: 37.4, current_wbgt_c: 39.2, grid_temp_c: 34.9,
    uhi_delta_c: 2.5, trigger_probability_7d: 0.92, model_tier: 'ensemble',
    consecutive_hot_days: 6, annual_trigger_days: 74,
  },
  {
    zone_id: 'DAR-MNZ', name: 'Manzese', settlement_type: 'informal', heat_vulnerability: 'severe',
    latitude: -6.79, longitude: 39.245, elevation_m: 34, worker_population_est: 14_820,
    outdoor_exposure_pct: 0.74, corrected_temp_c: 37.1, current_wbgt_c: 39.0, grid_temp_c: 34.7,
    uhi_delta_c: 2.4, trigger_probability_7d: 0.88, model_tier: 'ensemble',
    consecutive_hot_days: 5, annual_trigger_days: 71,
  },
  {
    zone_id: 'DAR-TAN', name: 'Tandale', settlement_type: 'informal', heat_vulnerability: 'severe',
    latitude: -6.785, longitude: 39.255, elevation_m: 32, worker_population_est: 12_640,
    outdoor_exposure_pct: 0.76, corrected_temp_c: 37.2, current_wbgt_c: 38.9, grid_temp_c: 34.6,
    uhi_delta_c: 2.6, trigger_probability_7d: 0.85, model_tier: 'ensemble',
    consecutive_hot_days: 5, annual_trigger_days: 68,
  },
  {
    zone_id: 'DAR-BUG', name: 'Buguruni', settlement_type: 'informal', heat_vulnerability: 'high',
    latitude: -6.82, longitude: 39.26, elevation_m: 21, worker_population_est: 8_720,
    outdoor_exposure_pct: 0.71, corrected_temp_c: 36.8, current_wbgt_c: 38.9, grid_temp_c: 34.3,
    uhi_delta_c: 2.5, trigger_probability_7d: 0.81, model_tier: 'full_model',
    consecutive_hot_days: 4, annual_trigger_days: 62,
  },
  {
    zone_id: 'DAR-VIN', name: 'Vingunguti', settlement_type: 'informal', heat_vulnerability: 'high',
    latitude: -6.835, longitude: 39.25, elevation_m: 24, worker_population_est: 9_840,
    outdoor_exposure_pct: 0.73, corrected_temp_c: 36.9, current_wbgt_c: 38.6, grid_temp_c: 34.4,
    uhi_delta_c: 2.5, trigger_probability_7d: 0.77, model_tier: 'full_model',
    consecutive_hot_days: 4, annual_trigger_days: 59,
  },
  {
    zone_id: 'DAR-JAN', name: 'Jangwani', settlement_type: 'informal', heat_vulnerability: 'severe',
    latitude: -6.8, longitude: 39.27, elevation_m: 6, worker_population_est: 7_310,
    outdoor_exposure_pct: 0.79, corrected_temp_c: 37.0, current_wbgt_c: 38.5, grid_temp_c: 34.1,
    uhi_delta_c: 2.9, trigger_probability_7d: 0.74, model_tier: 'ensemble',
    consecutive_hot_days: 3, annual_trigger_days: 83,
  },
  {
    zone_id: 'DAR-MBU', name: 'Mburahati', settlement_type: 'informal', heat_vulnerability: 'high',
    latitude: -6.805, longitude: 39.25, elevation_m: 29, worker_population_est: 6_520,
    outdoor_exposure_pct: 0.7, corrected_temp_c: 36.6, current_wbgt_c: 38.2, grid_temp_c: 34.1,
    uhi_delta_c: 2.5, trigger_probability_7d: 0.68, model_tier: 'full_model',
    consecutive_hot_days: 3, annual_trigger_days: 56,
  },

  // ── Mixed settlements
  {
    zone_id: 'DAR-KIG', name: 'Kigogo', settlement_type: 'mixed', heat_vulnerability: 'medium',
    latitude: -6.79, longitude: 39.265, elevation_m: 26, worker_population_est: 8_140,
    outdoor_exposure_pct: 0.58, corrected_temp_c: 35.9, current_wbgt_c: 37.9, grid_temp_c: 34.0,
    uhi_delta_c: 1.9, trigger_probability_7d: 0.54, model_tier: 'full_model',
    consecutive_hot_days: 2, annual_trigger_days: 42,
  },
  {
    zone_id: 'DAR-MAG', name: 'Magomeni', settlement_type: 'mixed', heat_vulnerability: 'medium',
    latitude: -6.8, longitude: 39.26, elevation_m: 18, worker_population_est: 6_980,
    outdoor_exposure_pct: 0.55, corrected_temp_c: 35.6, current_wbgt_c: 37.6, grid_temp_c: 33.8,
    uhi_delta_c: 1.8, trigger_probability_7d: 0.47, model_tier: 'full_model',
    consecutive_hot_days: 2, annual_trigger_days: 38,
  },
  {
    zone_id: 'DAR-MWA', name: 'Mwananyamala', settlement_type: 'mixed', heat_vulnerability: 'medium',
    latitude: -6.778, longitude: 39.255, elevation_m: 31, worker_population_est: 5_860,
    outdoor_exposure_pct: 0.53, corrected_temp_c: 35.4, current_wbgt_c: 37.3, grid_temp_c: 33.7,
    uhi_delta_c: 1.7, trigger_probability_7d: 0.41, model_tier: 'persistence',
    consecutive_hot_days: 1, annual_trigger_days: 34,
  },
  {
    zone_id: 'DAR-MSA', name: 'Msasani', settlement_type: 'mixed', heat_vulnerability: 'medium',
    latitude: -6.76, longitude: 39.265, elevation_m: 14, worker_population_est: 4_910,
    outdoor_exposure_pct: 0.49, corrected_temp_c: 34.9, current_wbgt_c: 36.8, grid_temp_c: 33.4,
    uhi_delta_c: 1.5, trigger_probability_7d: 0.35, model_tier: 'persistence',
    consecutive_hot_days: 1, annual_trigger_days: 27,
  },

  // ── Formal settlements — cooler, better infrastructure
  {
    zone_id: 'DAR-KIN', name: 'Kinondoni', settlement_type: 'formal', heat_vulnerability: 'low',
    latitude: -6.77, longitude: 39.24, elevation_m: 42, worker_population_est: 4_280,
    outdoor_exposure_pct: 0.34, corrected_temp_c: 33.8, current_wbgt_c: 35.9, grid_temp_c: 33.1,
    uhi_delta_c: 0.7, trigger_probability_7d: 0.18, model_tier: 'persistence',
    consecutive_hot_days: 0, annual_trigger_days: 12,
  },
  {
    zone_id: 'DAR-SIN', name: 'Sinza', settlement_type: 'formal', heat_vulnerability: 'low',
    latitude: -6.775, longitude: 39.225, elevation_m: 48, worker_population_est: 3_760,
    outdoor_exposure_pct: 0.31, corrected_temp_c: 33.5, current_wbgt_c: 35.6, grid_temp_c: 32.9,
    uhi_delta_c: 0.6, trigger_probability_7d: 0.14, model_tier: 'climatology',
    consecutive_hot_days: 0, annual_trigger_days: 9,
  },
  {
    zone_id: 'DAR-KAW', name: 'Kawe', settlement_type: 'formal', heat_vulnerability: 'low',
    latitude: -6.755, longitude: 39.235, elevation_m: 12, worker_population_est: 3_240,
    outdoor_exposure_pct: 0.29, corrected_temp_c: 33.1, current_wbgt_c: 35.2, grid_temp_c: 32.6,
    uhi_delta_c: 0.5, trigger_probability_7d: 0.11, model_tier: 'climatology',
    consecutive_hot_days: 0, annual_trigger_days: 6,
  },
  {
    zone_id: 'DAR-MIK', name: 'Mikocheni', settlement_type: 'formal', heat_vulnerability: 'low',
    latitude: -6.765, longitude: 39.27, elevation_m: 8, worker_population_est: 2_178,
    outdoor_exposure_pct: 0.26, corrected_temp_c: 32.7, current_wbgt_c: 34.8, grid_temp_c: 32.3,
    uhi_delta_c: 0.4, trigger_probability_7d: 0.08, model_tier: 'climatology',
    consecutive_hot_days: 0, annual_trigger_days: 4,
  },
]

function riskLevelFor(wbgt: number): string {
  if (wbgt > 38.8) return 'critical'
  if (wbgt > 37.5) return 'high'
  if (wbgt > 36) return 'moderate'
  return 'low'
}

// ── Build canonical Zone records ──────────────────────────────────────────

const TOTAL_WORKERS = ZONE_SEEDS.reduce((s, z) => s + z.worker_population_est, 0)

const ZONES: Zone[] = ZONE_SEEDS.map((z) => ({
  zone_id: z.zone_id,
  name: z.name,
  city: 'Dar es Salaam',
  country: 'Tanzania',
  latitude: z.latitude,
  longitude: z.longitude,
  elevation_m: z.elevation_m,
  settlement_type: z.settlement_type,
  worker_population_est: z.worker_population_est,
  outdoor_exposure_pct: z.outdoor_exposure_pct,
  heat_vulnerability: z.heat_vulnerability,
  risk_level: riskLevelFor(z.current_wbgt_c),
  corrected_temp_c: z.corrected_temp_c,
  current_wbgt_c: z.current_wbgt_c,
  current_heat_index_c: z.corrected_temp_c + 2.1,
  max_temp_c: z.corrected_temp_c + 1.4,
  max_wbgt_c: z.current_wbgt_c + 0.8,
  consecutive_hot_days: z.consecutive_hot_days,
  trigger_probability_7d: z.trigger_probability_7d,
  prediction_confidence: 0.72 + (z.model_tier === 'ensemble' ? 0.18 : z.model_tier === 'full_model' ? 0.12 : 0.04),
  model_tier: z.model_tier,
}))

export const ZONES_RESPONSE: ZonesResponse = {
  zones: ZONES,
  total: ZONES.length,
  cities: ['Dar es Salaam'],
}

// ── Daily heat indices (90 days per zone) ─────────────────────────────────

function buildDailyHistory(seed: ZoneSeed): DailyHeat[] {
  const rand = rng(seed.zone_id.split('').reduce((a, c) => a + c.charCodeAt(0), 0))
  const baseTemp = seed.corrected_temp_c - 1.2
  const baseWbgt = seed.current_wbgt_c - 1.0
  const out: DailyHeat[] = []
  for (let i = 89; i >= 0; i--) {
    // Light seasonality — April is peak, earlier days are cooler
    const seasonal = Math.sin(((89 - i) / 89) * Math.PI * 0.9) * 1.4
    const noise = (rand() - 0.5) * 1.4
    const temp = baseTemp + seasonal + noise
    const wbgt = baseWbgt + seasonal * 0.9 + noise * 0.7
    const gridTemp = temp - seed.uhi_delta_c - (rand() - 0.5) * 0.3
    out.push({
      date: dateOnly(i),
      temp_c: Math.round(temp * 10) / 10,
      grid_temp_c: Math.round(gridTemp * 10) / 10,
      wbgt_c: Math.round(wbgt * 10) / 10,
    })
  }
  return out
}

const INDICES: IndexData[] = ZONE_SEEDS.map((z) => ({
  zone_id: z.zone_id,
  zone_name: z.name,
  city: 'Dar es Salaam',
  temp_current: z.corrected_temp_c,
  wbgt_current: z.current_wbgt_c,
  risk_level: riskLevelFor(z.current_wbgt_c),
  daily_history: buildDailyHistory(z),
}))

export const INDICES_RESPONSE: IndicesResponse = {
  indices: INDICES,
  total: INDICES.length,
}

// ── Triggers (recent trigger events) ─────────────────────────────────────
// 28 events over the last ~120 days, concentrated in informal zones.

interface TriggerSeed {
  zone_id: string
  daysAgo: number
  level: 'critical' | 'high' | 'moderate'
  max_temp_c: number
  max_wbgt_c: number
  consecutive_days: number
  payout_per_worker_usd: number
}

const TRIGGER_SEEDS: TriggerSeed[] = [
  { zone_id: 'DAR-MAB', daysAgo: 0, level: 'critical', max_temp_c: 38.1, max_wbgt_c: 39.6, consecutive_days: 6, payout_per_worker_usd: 12 },
  { zone_id: 'DAR-MNZ', daysAgo: 0, level: 'critical', max_temp_c: 37.8, max_wbgt_c: 39.4, consecutive_days: 5, payout_per_worker_usd: 12 },
  { zone_id: 'DAR-TAN', daysAgo: 1, level: 'critical', max_temp_c: 37.6, max_wbgt_c: 39.2, consecutive_days: 5, payout_per_worker_usd: 12 },
  { zone_id: 'DAR-BUG', daysAgo: 1, level: 'high',     max_temp_c: 37.3, max_wbgt_c: 39.0, consecutive_days: 4, payout_per_worker_usd: 12 },
  { zone_id: 'DAR-VIN', daysAgo: 2, level: 'high',     max_temp_c: 37.4, max_wbgt_c: 38.8, consecutive_days: 4, payout_per_worker_usd: 12 },
  { zone_id: 'DAR-JAN', daysAgo: 3, level: 'critical', max_temp_c: 37.5, max_wbgt_c: 38.7, consecutive_days: 3, payout_per_worker_usd: 12 },
  { zone_id: 'DAR-MBU', daysAgo: 4, level: 'high',     max_temp_c: 36.9, max_wbgt_c: 38.4, consecutive_days: 3, payout_per_worker_usd: 12 },
  { zone_id: 'DAR-MAB', daysAgo: 11, level: 'critical', max_temp_c: 37.9, max_wbgt_c: 39.1, consecutive_days: 4, payout_per_worker_usd: 12 },
  { zone_id: 'DAR-MNZ', daysAgo: 12, level: 'critical', max_temp_c: 37.6, max_wbgt_c: 38.9, consecutive_days: 3, payout_per_worker_usd: 12 },
  { zone_id: 'DAR-BUG', daysAgo: 18, level: 'high',     max_temp_c: 36.8, max_wbgt_c: 38.7, consecutive_days: 3, payout_per_worker_usd: 12 },
  { zone_id: 'DAR-KIG', daysAgo: 22, level: 'moderate', max_temp_c: 36.2, max_wbgt_c: 38.2, consecutive_days: 2, payout_per_worker_usd: 8 },
  { zone_id: 'DAR-TAN', daysAgo: 27, level: 'critical', max_temp_c: 37.7, max_wbgt_c: 39.0, consecutive_days: 4, payout_per_worker_usd: 12 },
  { zone_id: 'DAR-MAG', daysAgo: 31, level: 'moderate', max_temp_c: 35.9, max_wbgt_c: 37.9, consecutive_days: 2, payout_per_worker_usd: 8 },
  { zone_id: 'DAR-VIN', daysAgo: 38, level: 'high',     max_temp_c: 37.1, max_wbgt_c: 38.6, consecutive_days: 3, payout_per_worker_usd: 12 },
  { zone_id: 'DAR-MAB', daysAgo: 44, level: 'critical', max_temp_c: 37.8, max_wbgt_c: 39.2, consecutive_days: 4, payout_per_worker_usd: 12 },
  { zone_id: 'DAR-MWA', daysAgo: 47, level: 'moderate', max_temp_c: 35.6, max_wbgt_c: 37.7, consecutive_days: 2, payout_per_worker_usd: 8 },
  { zone_id: 'DAR-MNZ', daysAgo: 52, level: 'critical', max_temp_c: 37.4, max_wbgt_c: 39.0, consecutive_days: 3, payout_per_worker_usd: 12 },
  { zone_id: 'DAR-BUG', daysAgo: 58, level: 'high',     max_temp_c: 36.9, max_wbgt_c: 38.8, consecutive_days: 3, payout_per_worker_usd: 12 },
  { zone_id: 'DAR-JAN', daysAgo: 64, level: 'critical', max_temp_c: 37.2, max_wbgt_c: 38.9, consecutive_days: 3, payout_per_worker_usd: 12 },
  { zone_id: 'DAR-MBU', daysAgo: 71, level: 'high',     max_temp_c: 36.7, max_wbgt_c: 38.5, consecutive_days: 2, payout_per_worker_usd: 12 },
  { zone_id: 'DAR-MAB', daysAgo: 78, level: 'critical', max_temp_c: 37.6, max_wbgt_c: 39.1, consecutive_days: 3, payout_per_worker_usd: 12 },
  { zone_id: 'DAR-TAN', daysAgo: 84, level: 'high',     max_temp_c: 36.9, max_wbgt_c: 38.7, consecutive_days: 2, payout_per_worker_usd: 12 },
  { zone_id: 'DAR-MSA', daysAgo: 89, level: 'moderate', max_temp_c: 35.2, max_wbgt_c: 37.4, consecutive_days: 2, payout_per_worker_usd: 8 },
  { zone_id: 'DAR-MNZ', daysAgo: 96, level: 'critical', max_temp_c: 37.5, max_wbgt_c: 39.0, consecutive_days: 4, payout_per_worker_usd: 12 },
  { zone_id: 'DAR-VIN', daysAgo: 102, level: 'high',    max_temp_c: 37.0, max_wbgt_c: 38.6, consecutive_days: 3, payout_per_worker_usd: 12 },
  { zone_id: 'DAR-MAB', daysAgo: 112, level: 'critical', max_temp_c: 37.7, max_wbgt_c: 39.2, consecutive_days: 4, payout_per_worker_usd: 12 },
  { zone_id: 'DAR-BUG', daysAgo: 118, level: 'high',    max_temp_c: 36.8, max_wbgt_c: 38.7, consecutive_days: 3, payout_per_worker_usd: 12 },
  { zone_id: 'DAR-JAN', daysAgo: 126, level: 'critical', max_temp_c: 37.3, max_wbgt_c: 38.9, consecutive_days: 3, payout_per_worker_usd: 12 },
]

const ZONE_BY_ID: Record<string, (typeof ZONE_SEEDS)[number]> = Object.fromEntries(
  ZONE_SEEDS.map((z) => [z.zone_id, z])
)

// Trigger events are sized to match the `trigger_events` table the real
// backend reads in `api/triggers.ts`. We don't materialise `enrolled_workers`
// or `total_payout_usd` here — the backend query doesn't return them, so
// keeping them out of the mock ensures the UI never grows a silent reliance.
const TRIGGERS: Trigger[] = TRIGGER_SEEDS.map((t) => {
  const z = ZONE_BY_ID[t.zone_id]
  return {
    zone_id: t.zone_id,
    zone_name: z.name,
    city: 'Dar es Salaam',
    trigger_level: t.level,
    triggered_at: isoDaysAgo(t.daysAgo, 9, 0),
    heat_risk_score: Math.round((t.max_wbgt_c - 32) * 18),
    max_temp_c: t.max_temp_c,
    max_wbgt_c: t.max_wbgt_c,
    consecutive_days: t.consecutive_days,
    settlement_type: z.settlement_type,
    payout_per_worker_usd: t.payout_per_worker_usd,
  }
})

export const TRIGGERS_RESPONSE: TriggersResponse = {
  triggers: TRIGGERS,
  total: TRIGGERS.length,
}

// Total worker-payouts this season = sum of enrolled × events.
const PAYOUTS_THIS_SEASON = TRIGGER_SEEDS.reduce(
  (s, t) => s + Math.round(ZONE_BY_ID[t.zone_id].worker_population_est * 0.35),
  0
)

// ── Basis-risk assessments ────────────────────────────────────────────────
// Shape matches `api/basis-risk.ts`: rows selected from the `basis_risk`
// table, joined to `zones` for name/city/settlement context.

const BASIS_RISK: BasisRisk[] = ZONE_SEEDS.map((z) => {
  // Informal zones are more vulnerable, so they have slightly higher
  // false-negative rates (more heat-days the trigger misses) but better
  // correlation overall because the signal is stronger.
  const fpr = z.settlement_type === 'formal' ? 0.12 : z.settlement_type === 'mixed' ? 0.09 : 0.07
  const fnr = z.settlement_type === 'informal' ? 0.11 : z.settlement_type === 'mixed' ? 0.08 : 0.05
  const correlation = z.settlement_type === 'informal' ? 0.78 : z.settlement_type === 'mixed' ? 0.71 : 0.64
  const mae = Math.round((1 - correlation) * 100) / 100
  return {
    zone_id: z.zone_id,
    zone_name: z.name,
    city: 'Dar es Salaam',
    settlement_type: z.settlement_type,
    false_positive_rate: fpr,
    false_negative_rate: fnr,
    correlation,
    mae,
    accuracy_by_tier: {
      persistence: 0.62,
      full_model: 0.74,
      ensemble: 0.82,
    },
    recommendations: null,
  }
})

export const BASIS_RISK_RESPONSE: BasisRiskResponse = {
  basis_risk: BASIS_RISK,
  total: BASIS_RISK.length,
}

// ── Notifications (worker-facing alerts dispatched by the pipeline) ───────

// Shape matches `api/notifications.ts`: rows from the `notifications`
// table joined to `zones` and `trigger_events`. `recipient_count` is
// hardcoded to 1 server-side; we mirror realistic fan-out here.
const NOTIFICATIONS: Notification[] = [
  { id: 'msg_3201', zone_id: 'DAR-MAB', zone_name: 'Mabibo', city: 'Dar es Salaam', trigger_level: 'critical',
    channel: 'SMS', language: 'sw', recipient_count: 4182, status: 'delivered',
    message_preview: 'Tahadhari: joto kali Mabibo. Payout TSh 30,000 imetumwa kwa M-Pesa.',
    delivered_at: isoDaysAgo(0, 9, 12) },
  { id: 'msg_3202', zone_id: 'DAR-MNZ', zone_name: 'Manzese', city: 'Dar es Salaam', trigger_level: 'critical',
    channel: 'SMS', language: 'sw', recipient_count: 5187, status: 'delivered',
    message_preview: 'Tahadhari: joto kali Manzese. Payout TSh 30,000 imetumwa kwa M-Pesa.',
    delivered_at: isoDaysAgo(0, 9, 14) },
  { id: 'msg_3203', zone_id: 'DAR-TAN', zone_name: 'Tandale', city: 'Dar es Salaam', trigger_level: 'critical',
    channel: 'SMS', language: 'sw', recipient_count: 4424, status: 'delivered',
    message_preview: 'Tahadhari: joto kali Tandale. Payout TSh 30,000 imetumwa kwa M-Pesa.',
    delivered_at: isoDaysAgo(1, 9, 10) },
  { id: 'msg_3204', zone_id: 'DAR-BUG', zone_name: 'Buguruni', city: 'Dar es Salaam', trigger_level: 'high',
    channel: 'SMS', language: 'sw', recipient_count: 3052, status: 'delivered',
    message_preview: 'Onyo: joto juu Buguruni. Fuata kanuni za usalama. Payout TSh 30,000.',
    delivered_at: isoDaysAgo(1, 9, 15) },
  { id: 'msg_3205', zone_id: 'DAR-VIN', zone_name: 'Vingunguti', city: 'Dar es Salaam', trigger_level: 'high',
    channel: 'SMS', language: 'sw', recipient_count: 3444, status: 'delivered',
    message_preview: 'Onyo: joto juu Vingunguti. Payout TSh 30,000 imetumwa kwa M-Pesa.',
    delivered_at: isoDaysAgo(2, 9, 8) },
  { id: 'msg_3206', zone_id: 'DAR-JAN', zone_name: 'Jangwani', city: 'Dar es Salaam', trigger_level: 'critical',
    channel: 'SMS', language: 'sw', recipient_count: 2559, status: 'delivered',
    message_preview: 'Tahadhari: joto kali Jangwani. Payout TSh 30,000 imetumwa kwa M-Pesa.',
    delivered_at: isoDaysAgo(3, 9, 22) },
  { id: 'msg_3207', zone_id: 'DAR-MBU', zone_name: 'Mburahati', city: 'Dar es Salaam', trigger_level: 'high',
    channel: 'SMS', language: 'sw', recipient_count: 2282, status: 'delivered',
    message_preview: 'Onyo: joto juu Mburahati. Pigeni simu 199 kwa usaidizi.',
    delivered_at: isoDaysAgo(4, 9, 18) },
]

const NOTIFICATIONS_BY_LANGUAGE: Record<string, number> = NOTIFICATIONS.reduce(
  (acc, n) => {
    acc[n.language] = (acc[n.language] ?? 0) + 1
    return acc
  },
  {} as Record<string, number>
)

export const NOTIFICATIONS_RESPONSE: NotificationsResponse = {
  notifications: NOTIFICATIONS,
  by_language: NOTIFICATIONS_BY_LANGUAGE,
}

// ── Disbursements ─────────────────────────────────────────────────────────
// Shape matches `src/types.ts::DisbursementsResponse`. One row per zone,
// describing the current-week payout posture. Zones above the city-wide
// P97 WBGT (38.8°C) are flagged `triggers_payout`.

const CITY_P97_WBGT = 38.8
const TZS_PER_USD = 2_500
const PROVIDER_CYCLE = ['M-Pesa', 'Tigo Pesa', 'Airtel Money'] as const

const DISBURSEMENT_RECORDS: DisbursementRecord[] = ZONE_SEEDS.map((z, idx) => {
  const triggers = z.current_wbgt_c > CITY_P97_WBGT
  const enrolled = Math.round(z.worker_population_est * 0.35)
  // Payouts are higher in informal zones where workers face the worst
  // conditions and have the least fallback income.
  const payoutUsd = z.settlement_type === 'informal' ? 12 : z.settlement_type === 'mixed' ? 8 : 6
  const payoutTzs = payoutUsd * TZS_PER_USD
  // Historical disbursements: more events for hotter zones, but not round.
  const pastEvents = Math.round(z.annual_trigger_days / 8)
  const pastTotalUsd = pastEvents * enrolled * payoutUsd * 0.92
  return {
    zone_id: z.zone_id,
    zone_name: z.name,
    city: 'Dar es Salaam',
    settlement_type: z.settlement_type,
    heat_vulnerability: z.heat_vulnerability,
    current_wbgt: z.current_wbgt_c,
    threshold: CITY_P97_WBGT,
    triggers_payout: triggers,
    workers: enrolled,
    payout_per_worker_usd: payoutUsd,
    payout_per_worker_tzs: payoutTzs,
    zone_total_usd: triggers ? enrolled * payoutUsd : 0,
    zone_total_tzs: triggers ? enrolled * payoutTzs : 0,
    provider: PROVIDER_CYCLE[idx % PROVIDER_CYCLE.length],
    status: triggers ? 'pending' : 'no_trigger',
    past_disbursements: pastEvents,
    past_total_usd: Math.round(pastTotalUsd),
  }
})

const TRIGGERING_NOW = DISBURSEMENT_RECORDS.filter((d) => d.triggers_payout)

export const DISBURSEMENTS_RESPONSE: DisbursementsResponse = {
  disbursements: DISBURSEMENT_RECORDS,
  stats: {
    total_usd: TRIGGERING_NOW.reduce((s, d) => s + d.zone_total_usd, 0),
    total_tzs: TRIGGERING_NOW.reduce((s, d) => s + d.zone_total_tzs, 0),
    total_workers_paid: TRIGGERING_NOW.reduce((s, d) => s + d.workers, 0),
    zones_paying: TRIGGERING_NOW.length,
    zones_total: DISBURSEMENT_RECORDS.length,
    threshold_wbgt: CITY_P97_WBGT,
  },
}

// ── Pipeline runs + stats ────────────────────────────────────────────────

interface RunSeed {
  daysAgo: number
  status: 'success' | 'partial' | 'failed'
  zones_processed: number
  triggers_found: number
  notifications_sent: number
  total_cost_usd: number
  duration_s: number
  note?: string
}

const RUN_SEEDS: RunSeed[] = [
  { daysAgo: 0, status: 'success',  zones_processed: 15, triggers_found: 7, notifications_sent: 25130, total_cost_usd: 0.1842, duration_s: 412 },
  { daysAgo: 1, status: 'success',  zones_processed: 15, triggers_found: 4, notifications_sent: 15102, total_cost_usd: 0.1531, duration_s: 388 },
  { daysAgo: 2, status: 'success',  zones_processed: 15, triggers_found: 3, notifications_sent: 12440, total_cost_usd: 0.1482, duration_s: 401 },
  { daysAgo: 3, status: 'success',  zones_processed: 15, triggers_found: 2, notifications_sent: 9284,  total_cost_usd: 0.1394, duration_s: 376 },
  { daysAgo: 4, status: 'partial',  zones_processed: 15, triggers_found: 5, notifications_sent: 18421, total_cost_usd: 0.1612, duration_s: 442,
    note: 'M-Pesa B2C rate-limited 312 transfers; retried successfully' },
  { daysAgo: 5, status: 'success',  zones_processed: 15, triggers_found: 1, notifications_sent: 4182,  total_cost_usd: 0.1328, duration_s: 369 },
  { daysAgo: 6, status: 'success',  zones_processed: 15, triggers_found: 0, notifications_sent: 0,     total_cost_usd: 0.1284, duration_s: 372 },
  { daysAgo: 7, status: 'success',  zones_processed: 15, triggers_found: 2, notifications_sent: 8112,  total_cost_usd: 0.1366, duration_s: 384 },
  { daysAgo: 8, status: 'success',  zones_processed: 15, triggers_found: 1, notifications_sent: 3428,  total_cost_usd: 0.1298, duration_s: 358 },
  { daysAgo: 9, status: 'success',  zones_processed: 15, triggers_found: 2, notifications_sent: 7860,  total_cost_usd: 0.1341, duration_s: 391 },
  { daysAgo: 10, status: 'failed',  zones_processed: 4,  triggers_found: 0, notifications_sent: 0,     total_cost_usd: 0.0226, duration_s: 96,
    note: 'NASA POWER timeout — retry scheduled' },
  { daysAgo: 11, status: 'success', zones_processed: 15, triggers_found: 3, notifications_sent: 11248, total_cost_usd: 0.1407, duration_s: 380 },
  { daysAgo: 12, status: 'success', zones_processed: 15, triggers_found: 1, notifications_sent: 3860,  total_cost_usd: 0.1288, duration_s: 365 },
  { daysAgo: 13, status: 'success', zones_processed: 15, triggers_found: 2, notifications_sent: 6964,  total_cost_usd: 0.1349, duration_s: 379 },
  { daysAgo: 14, status: 'success', zones_processed: 15, triggers_found: 0, notifications_sent: 0,     total_cost_usd: 0.1271, duration_s: 371 },
]

function buildSteps(seed: RunSeed): PipelineStep[] {
  const names: [string, number][] = [
    ['ingest', 48], ['heal', 72], ['index', 61],
    ['calibrate', 94], ['explain', 82], ['notify', 51],
  ]
  return names.map(([step, share]) => {
    // Failed runs stop at the step that broke.
    if (seed.status === 'failed' && step !== 'ingest') {
      return { step, status: step === 'heal' ? 'failed' : 'skipped', duration_s: 0 }
    }
    const frac = share / names.reduce((s, [, v]) => s + v, 0)
    const stepStatus = seed.status === 'partial' && step === 'notify' ? 'partial' : 'success'
    return { step, status: stepStatus, duration_s: Math.round(seed.duration_s * frac * 10) / 10 }
  })
}

const PIPELINE_RUNS: PipelineRun[] = RUN_SEEDS.map((seed) => {
  const startedAt = isoDaysAgo(seed.daysAgo, 2, 0)
  const endedAt = isoDaysAgo(seed.daysAgo, 2, Math.floor(seed.duration_s / 60))
  return {
    run_id: `run_${dateOnly(seed.daysAgo).replace(/-/g, '')}`,
    started_at: startedAt,
    ended_at: endedAt,
    status: seed.status === 'success' ? 'success' : seed.status,
    duration_s: seed.duration_s,
    zones_processed: seed.zones_processed,
    triggers_found: seed.triggers_found,
    notifications_sent: seed.notifications_sent,
    total_cost_usd: seed.total_cost_usd,
    steps: buildSteps(seed),
  }
})

export const PIPELINE_RUNS_RESPONSE: PipelineRunsResponse = {
  runs: PIPELINE_RUNS,
  total: PIPELINE_RUNS.length,
}

const SUCCESSFUL_RUNS = RUN_SEEDS.filter((r) => r.status === 'success').length
const TOTAL_COST = Math.round(RUN_SEEDS.reduce((s, r) => s + r.total_cost_usd, 0) * 10000) / 10000
const TOTAL_ENROLLED = ZONE_SEEDS.reduce((s, z) => s + Math.round(z.worker_population_est * 0.35), 0)

export const PIPELINE_STATS_RESPONSE: PipelineStats = {
  total_runs: RUN_SEEDS.length,
  successful_runs: SUCCESSFUL_RUNS,
  success_rate: Math.round((SUCCESSFUL_RUNS / RUN_SEEDS.length) * 100) / 100,
  zones_monitored: ZONES.length,
  cities: 1,
  active_triggers: TRIGGERING_NOW.length,
  total_enrolled: TOTAL_ENROLLED,
  total_cost_usd: TOTAL_COST,
  avg_cost_per_run_usd: Math.round((TOTAL_COST / RUN_SEEDS.length) * 10000) / 10000,
  last_run: isoDaysAgo(0, 2, 7),
  data_sources: ['ERA5-Land', 'NASA POWER', 'Tanzania Met'],
}

// ── Enrolled workers ─────────────────────────────────────────────────────

export const ENROLLED_RESPONSE: EnrolledResponse = {
  by_zone: ZONE_SEEDS.map((z) => ({
    zone_id: z.zone_id,
    enrolled: Math.round(z.worker_population_est * 0.35),
    worker_population: z.worker_population_est,
  })),
  total_enrolled: TOTAL_ENROLLED,
}

// ── Coverage recommendation (ProgramDesigner page) ───────────────────────
// Shape matches `/api/coverage-recommendation`. Returns one record per
// filter combination — we derive from ZONE_SEEDS on the fly so the filter
// buttons in the UI actually do something.

interface ZonePopulation {
  total_pop: number
  outdoor_informal: number
  pct_women: number
}

const ZONE_POPULATIONS: Record<string, ZonePopulation> = {
  'DAR-MAB': { total_pop: 120000, outdoor_informal: 11950, pct_women: 0.62 },
  'DAR-MNZ': { total_pop: 150000, outdoor_informal: 14820, pct_women: 0.61 },
  'DAR-TAN': { total_pop: 128000, outdoor_informal: 12640, pct_women: 0.6 },
  'DAR-BUG': { total_pop: 88000,  outdoor_informal: 8720,  pct_women: 0.56 },
  'DAR-VIN': { total_pop: 102000, outdoor_informal: 9840,  pct_women: 0.57 },
  'DAR-JAN': { total_pop: 78000,  outdoor_informal: 7310,  pct_women: 0.58 },
  'DAR-MBU': { total_pop: 66000,  outdoor_informal: 6520,  pct_women: 0.55 },
  'DAR-KIG': { total_pop: 82000,  outdoor_informal: 8140,  pct_women: 0.54 },
  'DAR-MAG': { total_pop: 71000,  outdoor_informal: 6980,  pct_women: 0.55 },
  'DAR-MWA': { total_pop: 59000,  outdoor_informal: 5860,  pct_women: 0.53 },
  'DAR-MSA': { total_pop: 49000,  outdoor_informal: 4910,  pct_women: 0.52 },
  'DAR-KIN': { total_pop: 108000, outdoor_informal: 4280,  pct_women: 0.5 },
  'DAR-SIN': { total_pop: 95000,  outdoor_informal: 3760,  pct_women: 0.49 },
  'DAR-KAW': { total_pop: 82000,  outdoor_informal: 3240,  pct_women: 0.48 },
  'DAR-MIK': { total_pop: 55000,  outdoor_informal: 2178,  pct_women: 0.47 },
}

type Gender = 'all' | 'women' | 'men'
type SettlementFilter = 'all' | 'informal' | 'mixed' | 'formal'

export function buildCoverage(
  _payoutUsd: number,
  gender: Gender,
  settlementFilter: SettlementFilter
): CoverageResponse {
  const scenarioLabel = [
    gender === 'all' ? 'All' : gender === 'women' ? 'Women' : 'Men',
    settlementFilter === 'all' ? 'outdoor workers' : `in ${settlementFilter} settlements`,
  ].join(' ')

  let totalWorkers = 0
  let weeklyCostTotal = 0
  let annualCostTotal = 0
  let zonesTriggering = 0

  const zoneRecs: CoverageZone[] = []
  for (const z of ZONE_SEEDS) {
    if (settlementFilter !== 'all' && z.settlement_type !== settlementFilter) continue
    const pop = ZONE_POPULATIONS[z.zone_id]
    if (!pop) continue
    let enrolled = pop.outdoor_informal
    if (gender === 'women') enrolled = Math.floor(enrolled * pop.pct_women)
    if (gender === 'men') enrolled = Math.floor(enrolled * (1 - pop.pct_women))
    if (enrolled === 0) continue

    const payoutTriggered = z.current_wbgt_c > CITY_P97_WBGT
    const pf = z.trigger_probability_7d
    const alertPayout = Math.round((2 + pf * 3) * 100) / 100
    const insurancePayout = Math.round((7 + pf * 13) * 100) / 100
    const weeklyCostPerWorker = payoutTriggered ? insurancePayout : 0
    // Annual cost = insurance_payout × (trigger days / 7), matching the
    // real handler's arithmetic so rec-level totals look believable.
    const annualTriggerDays = Math.min(z.annual_trigger_days, 90)
    const annualCostPerWorker = Math.round(insurancePayout * (annualTriggerDays / 7) * 100) / 100
    const annualWorkerShare = 3
    const annualPhilanthropy = Math.round(annualCostPerWorker * 0.45 * 100) / 100
    const annualInsurer = Math.round(Math.max(0, annualCostPerWorker - annualWorkerShare - annualPhilanthropy) * 100) / 100

    if (payoutTriggered) zonesTriggering++
    totalWorkers += enrolled
    weeklyCostTotal += weeklyCostPerWorker * enrolled
    annualCostTotal += annualCostPerWorker * enrolled

    const urgency = pf > 0.8 ? 'critical' : pf > 0.5 ? 'high' : pf > 0.3 ? 'moderate' : 'low'

    zoneRecs.push({
      zone_id: z.zone_id,
      zone_name: z.name,
      city: 'Dar es Salaam',
      settlement_type: z.settlement_type,
      heat_vulnerability: z.heat_vulnerability,
      urgency,
      current_temp_c: Math.round(z.corrected_temp_c * 10) / 10,
      current_wbgt_c: Math.round(z.current_wbgt_c * 10) / 10,
      trigger_probability_7d: pf,
      triggers_this_week: payoutTriggered,
      risk_level: urgency,
      outdoor_exposure_pct: z.outdoor_exposure_pct,
      enrolled_workers: enrolled,
      weekly_cost_per_worker: Math.round(weeklyCostPerWorker * 100) / 100,
      alert_payout: payoutTriggered ? alertPayout : 0,
      insurance_payout: payoutTriggered && pf > 0.8 ? insurancePayout : 0,
      total_payout_this_week: Math.round(weeklyCostPerWorker * 100) / 100,
      worker_contribution: Math.round((3 / 52) * 100) / 100,
      philanthropy_share: 0,
      insurer_premium: weeklyCostPerWorker,
      neural_model: z.model_tier === 'ensemble' || z.model_tier === 'full_model',
      learned_frequency: z.annual_trigger_days,
      payout_factor: pf,
      annual_cost_per_worker: annualCostPerWorker,
      annual_worker_share: annualWorkerShare,
      annual_philanthropy_share: annualPhilanthropy,
      annual_insurer_share: annualInsurer,
      annual_trigger_days: annualTriggerDays,
    })
  }

  return {
    recommendation: {
      weekly_cost_per_worker:
        totalWorkers > 0 ? Math.round((weeklyCostTotal / totalWorkers) * 100) / 100 : 0,
      weekly_cost_total: Math.round(weeklyCostTotal),
      annual_cost_total: Math.round(annualCostTotal),
      annual_cost_per_worker:
        totalWorkers > 0 ? Math.round((annualCostTotal / totalWorkers) * 100) / 100 : 0,
      total_workers: totalWorkers,
      zones_triggering: zonesTriggering,
      zones_total: zoneRecs.length,
      model_type: 'neural_evt',
      scenario: scenarioLabel,
      payout_threshold_wbgt: CITY_P97_WBGT,
    },
    zones: zoneRecs,
    cost_summary: {
      total_to_workers_pct: 85,
      total_admin_pct: 7,
      total_basis_risk_pct: 8,
      worker_contribution_pct: 15,
      philanthropy_pct: 45,
      insurer_pct: 40,
    },
  }
}

// ── Calibrate response (used by an optional calibrate tool; kept here so
// the hook has real data if anyone wires it up) ──────────────────────────

export function buildCalibrate(params: CalibrateParams): CalibrateResponse {
  const zones: CalibrateZoneResult[] = ZONE_SEEDS.map((z, idx) => {
    const enrolled = Math.round(z.worker_population_est * 0.35)
    const daysAboveTemp = Math.round(z.annual_trigger_days * 1.1)
    const daysAboveWbgt = z.annual_trigger_days
    const events = Math.max(1, Math.round(z.annual_trigger_days / 3.4))
    const eventsPerYear = events
    const annualPayoutPerWorker = Math.round(events * params.payout_usd * 100) / 100
    const annualPayoutTotal = Math.round(annualPayoutPerWorker * enrolled)
    return {
      zone_id: z.zone_id,
      zone_name: z.name,
      city: 'Dar es Salaam',
      settlement_type: z.settlement_type,
      heat_vulnerability: z.heat_vulnerability,
      enrolled_workers: enrolled,
      days_above_temp: daysAboveTemp,
      days_above_wbgt: daysAboveWbgt,
      consecutive_days_temp: z.consecutive_hot_days + 1,
      consecutive_days_wbgt: z.consecutive_hot_days,
      trigger_events: events,
      events_per_year: eventsPerYear,
      annual_payout_per_worker: annualPayoutPerWorker,
      annual_payout_total: annualPayoutTotal,
      basis_risk_score: Math.round((0.74 - (z.settlement_type === 'formal' ? 0.12 : 0) - (z.settlement_type === 'mixed' ? 0.06 : 0)) * 100) / 100,
      triggered: z.current_wbgt_c > params.wbgt_threshold,
      actuarial_cost_per_worker: Math.round(annualPayoutPerWorker * 1.17 * 100) / 100,
      cost_breakdown: { payouts: annualPayoutTotal, admin: Math.round(annualPayoutTotal * 0.08), reserves: Math.round(annualPayoutTotal * 0.09) },
      allocated_budget: Math.round(annualPayoutTotal * 1.17),
      workers_covered: enrolled,
      coverage_pct: 1,
      priority_rank: idx + 1,
    }
  })

  const totalAnnualCost = zones.reduce((s, z) => s + z.annual_payout_total, 0)
  return {
    zones,
    summary: {
      total_zones: zones.length,
      zones_triggered: zones.filter((z) => z.triggered).length,
      total_trigger_days: zones.reduce((s, z) => s + z.days_above_wbgt, 0),
      avg_events_per_year: Math.round(
        (zones.reduce((s, z) => s + z.events_per_year, 0) / zones.length) * 10
      ) / 10,
      total_annual_cost: totalAnnualCost,
      avg_cost_per_worker:
        Math.round(
          (zones.reduce((s, z) => s + z.annual_payout_per_worker, 0) / zones.length) * 100
        ) / 100,
      total_enrolled: TOTAL_ENROLLED,
      avg_basis_risk:
        Math.round(
          (zones.reduce((s, z) => s + z.basis_risk_score, 0) / zones.length) * 100
        ) / 100,
    },
    allocation: {
      budget_usd: params.budget_usd,
      worker_contribution_usd: params.worker_contribution_usd,
      workers_covered: TOTAL_ENROLLED,
      overall_coverage_pct: 1,
      zones_fully_funded: zones.length,
      zones_partially_funded: 0,
      zones_unfunded: 0,
      stretch_analysis: {
        additional_workers_coverable: 18_400,
        stretch_budget_usd: Math.round(params.budget_usd * 1.25),
      },
    },
    thresholds: params,
  }
}

// ── Derived exports the UI uses for hero / metric consistency ────────────
//
// These are not wired into the API contract — they're available for any
// component that needs season-level roll-ups derived from the same fixture.

export const METRICS = {
  total_workers: TOTAL_WORKERS,
  total_enrolled: TOTAL_ENROLLED,
  zones_monitored: ZONES.length,
  payouts_this_season: PAYOUTS_THIS_SEASON,
  annual_cost_usd: 2_680_000,
  average_payout_usd: 10,
}