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
// Worker configuration, read once at boot from the environment. Every knob has
// a production default, so the Cloud Run service needs only credentials.
function int(env, name, fallback, { min = 0, max = Number.MAX_SAFE_INTEGER } = {}) {
const raw = env[name];
if (raw == null || raw === '') return fallback;
const n = Number(raw);
if (!Number.isFinite(n)) throw new Error(`${name} must be a number, got "${raw}"`);
return Math.min(max, Math.max(min, Math.floor(n)));
}
function list(env, name) {
const raw = String(env[name] || '').trim();
return raw ? raw.split(',').map((s) => s.trim()).filter(Boolean) : null;
}
/**
* @typedef {object} GatewayConfig
* @property {number} port health endpoint port (Cloud Run injects PORT)
* @property {number} concurrency inbox rows processed at once (one per chat)
* @property {number} pollMs idle wait between claims when the inbox is empty
* @property {number} leaseSeconds how long a claimed row is held before another worker may take it
* @property {number} maxAttempts attempts before a row is dead-lettered
* @property {number} backoffBaseMs first retry delay; doubles per attempt
* @property {number} backoffMaxMs ceiling on a single retry delay
* @property {number} turnTimeoutMs a row still running after this is failed and its lease released
* @property {number} sweepMs preview expiry sweep and inbox prune interval
* @property {number} pruneKeepDays finished rows older than this are deleted
* @property {number} shutdownGraceMs how long SIGTERM waits for in-flight rows
* @property {string[]|null} platforms restrict the adapters that load (default: every configured one)
* @property {string[]|null} chatKeys drain only these chats (debugging one chat)
* @property {boolean} telegramPolling pull Telegram updates with getUpdates instead of the webhook (local development)
* @property {boolean} discordGateway hold a Discord gateway connection for DMs and mentions
*/
/** @returns {GatewayConfig} */
export function loadConfig(env = process.env) {
return {
port: int(env, 'PORT', 8080, { min: 1, max: 65535 }),
concurrency: int(env, 'GATEWAY_CONCURRENCY', 8, { min: 1, max: 64 }),
pollMs: int(env, 'GATEWAY_POLL_MS', 1500, { min: 100 }),
leaseSeconds: int(env, 'GATEWAY_LEASE_SECONDS', 90, { min: 15 }),
maxAttempts: int(env, 'GATEWAY_MAX_ATTEMPTS', 5, { min: 1, max: 20 }),
backoffBaseMs: int(env, 'GATEWAY_BACKOFF_BASE_MS', 5000, { min: 0 }),
backoffMaxMs: int(env, 'GATEWAY_BACKOFF_MAX_MS', 300_000, { min: 0 }),
turnTimeoutMs: int(env, 'GATEWAY_TURN_TIMEOUT_MS', 300_000, { min: 1000 }),
sweepMs: int(env, 'GATEWAY_SWEEP_MS', 60_000, { min: 1000 }),
pruneKeepDays: int(env, 'GATEWAY_PRUNE_KEEP_DAYS', 7, { min: 1 }),
shutdownGraceMs: int(env, 'GATEWAY_SHUTDOWN_GRACE_MS', 8000, { min: 0 }),
platforms: list(env, 'GATEWAY_PLATFORMS'),
chatKeys: list(env, 'GATEWAY_CHAT_KEYS'),
telegramPolling: env.GATEWAY_TELEGRAM_POLLING === '1',
discordGateway: env.GATEWAY_DISCORD_GATEWAY !== '0',
};
}