Open-source platform for 3D AI agents. Turn text or a photo into a rigged, animated GLB avatar, give it an LLM brain, memory and a wallet, and embed it anywhere with one web component. Remote MCP server, x402 payments, Three.js, browser-native.
JavaScript 3d-generationai-agentsanimationavatardigital-humangenerative-aiglbgltfimage-to-3dllmmcpmcp-servermodel-context-protocolsolanatext-to-3dthreejsweb-componentswebglwebxrx402
// Temporal Anomaly Detection Service
// Detects when current activity levels deviate from historical baselines
// Backed by Upstash Redis via /api/temporal-baseline edge function
export type TemporalEventType =
| 'military_flights'
| 'tracked_flights'
| 'vessels'
| 'protests'
| 'news'
| 'ais_gaps'
| 'satellite_fires';
export interface TemporalAnomaly {
type: TemporalEventType;
region: string;
currentCount: number;
expectedCount: number;
zScore: number;
message: string;
severity: 'medium' | 'high' | 'critical';
}
const BASELINE_API = '/api/temporal-baseline';
const TYPE_LABELS: Record<TemporalEventType, string> = {
military_flights: 'Whale transactions',
tracked_flights: 'Whale transactions',
vessels: 'Protocol activity',
protests: 'Governance proposals',
news: 'News velocity',
ais_gaps: 'Dark wallet activity',
satellite_fires: 'On-chain anomaly detections',
};
const WEEKDAY_NAMES = ['Sunday', 'Monday', 'Tuesday', 'Wednesday', 'Thursday', 'Friday', 'Saturday'];
const MONTH_NAMES = ['', 'January', 'February', 'March', 'April', 'May', 'June',
'July', 'August', 'September', 'October', 'November', 'December'];
function formatAnomalyMessage(
type: TemporalEventType,
_region: string,
count: number,
mean: number,
multiplier: number,
): string {
const now = new Date();
const weekday = WEEKDAY_NAMES[now.getUTCDay()];
const month = MONTH_NAMES[now.getUTCMonth() + 1];
const mult = multiplier < 10 ? `${multiplier.toFixed(1)}x` : `${Math.round(multiplier)}x`;
return `${TYPE_LABELS[type]} ${mult} normal for ${weekday} (${month}) - ${count} vs baseline ${Math.round(mean)}`;
}
function getSeverity(zScore: number): 'medium' | 'high' | 'critical' {
if (zScore >= 3.0) return 'critical';
if (zScore >= 2.0) return 'high';
return 'medium';
}
// Fire-and-forget baseline update
export async function reportMetrics(
updates: Array<{ type: TemporalEventType; region: string; count: number }>
): Promise<void> {
try {
await fetch(BASELINE_API, {
method: 'POST',
headers: { 'Content-Type': 'application/json' },
body: JSON.stringify({ updates }),
});
} catch (e) {
console.warn('[TemporalBaseline] Update failed:', e);
}
}
// Check for anomaly (returns null if learning or normal)
export async function checkAnomaly(
type: TemporalEventType,
region: string,
count: number,
): Promise<TemporalAnomaly | null> {
try {
const params = new URLSearchParams({ type, region, count: String(count) });
const res = await fetch(`${BASELINE_API}?${params}`);
if (!res.ok) return null;
const data = await res.json();
if (!data.anomaly) return null;
return {
type,
region,
currentCount: count,
expectedCount: Math.round(data.baseline.mean),
zScore: data.anomaly.zScore,
severity: getSeverity(data.anomaly.zScore),
message: formatAnomalyMessage(type, region, count, data.baseline.mean, data.anomaly.multiplier),
};
} catch (e) {
console.warn('[TemporalBaseline] Check failed:', e);
return null;
}
}
// Batch: report metrics AND check for anomalies in one flow
export async function updateAndCheck(
metrics: Array<{ type: TemporalEventType; region: string; count: number }>
): Promise<TemporalAnomaly[]> {
// Fire-and-forget the update
reportMetrics(metrics).catch(() => {});
// Check anomalies in parallel
const results = await Promise.allSettled(
metrics.map(m => checkAnomaly(m.type, m.region, m.count))
);
return results
.filter((r): r is PromiseFulfilledResult<TemporalAnomaly | null> => r.status === 'fulfilled')
.map(r => r.value)
.filter((a): a is TemporalAnomaly => a !== null)
.sort((a, b) => b.zScore - a.zScore);
}