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
/**
* getTrendingSlim (api/_lib/pump-trending.js) — the shared trending feed
* behind /api/pump/trending and /api/community/worlds.
*
* Pins the contract that fixed the 2026-07-26 worlds outage: every upstream
* fetch asks for the canonical batch (FETCH_LIMIT) rather than the caller's
* limit, so ONE cache entry serves every consumer and the rate-limited
* Birdeye key is hit at most once per TTL. Before that, small-limit callers
* kept the cache too small for the 24-row communities caller, whose own
* burst of fetches 429'd Birdeye, tripped the breaker, and 503'd the route.
*/
import { describe, it, expect, vi, beforeEach } from 'vitest';
vi.mock('../api/_lib/db.js', () => ({
sql: vi.fn(async () => []),
}));
const BIRDEYE_ROWS = 20;
function birdeyeResponse(count) {
return {
ok: true,
json: async () => ({
data: {
tokens: Array.from({ length: count }, (_, i) => ({
address: 'Mint'.padEnd(40, String(i % 10)) + i,
symbol: `T${i}`,
name: `Token ${i}`,
logoURI: null,
price: 1 + i,
rank: i + 1,
})),
},
}),
};
}
async function freshModule() {
vi.resetModules();
process.env.BIRDEYE_API_KEY = 'test-key';
return import('../api/_lib/pump-trending.js');
}
beforeEach(() => {
vi.unstubAllGlobals();
});
describe('getTrendingSlim — one canonical fetch serves every caller', () => {
it('fetches the canonical batch once and slices it for different limits', async () => {
const calls = [];
vi.stubGlobal('fetch', vi.fn(async (url) => {
calls.push(String(url));
return birdeyeResponse(BIRDEYE_ROWS);
}));
const { getTrendingSlim } = await freshModule();
const small = await getTrendingSlim(3);
expect(small.data).toHaveLength(3);
// The 24-row caller (communities) must be served from the SAME cache
// entry: no second upstream call, no 503.
const large = await getTrendingSlim(24);
expect(large.data).toHaveLength(BIRDEYE_ROWS);
expect(large.stale).toBe(false);
expect(calls).toHaveLength(1);
expect(calls[0]).toContain('limit=20');
});
it('a large-limit caller still gets the full available feed on a cold cache', async () => {
vi.stubGlobal('fetch', vi.fn(async () => birdeyeResponse(BIRDEYE_ROWS)));
const { getTrendingSlim } = await freshModule();
const { data } = await getTrendingSlim(24);
expect(data).toHaveLength(BIRDEYE_ROWS);
});
it('serves a shorter stale cache instead of an outage when every rung fails', async () => {
let fail = false;
vi.stubGlobal('fetch', vi.fn(async () => {
if (fail) throw new Error('down');
return birdeyeResponse(BIRDEYE_ROWS);
}));
const { getTrendingSlim } = await freshModule();
await getTrendingSlim(3); // warm the cache
fail = true;
vi.useFakeTimers();
try {
// Past the 30s TTL but inside the 10 min stale window.
vi.setSystemTime(Date.now() + 60_000);
const { data, stale } = await getTrendingSlim(24);
expect(stale).toBe(true);
expect(data).toHaveLength(BIRDEYE_ROWS);
} finally {
vi.useRealTimers();
}
});
});