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
import 'dotenv/config'
import { writeFileSync } from 'node:fs'
import { fileURLToPath } from 'node:url'
import { serve } from '@hono/node-server'
import { createApp } from '../src/app.js'
/**
* Captures real responses from a locally-running instance of this server and
* writes them to `docs/samples.json`. The docs site embeds these as a
* clearly-labeled "sample of live output — captured <timestamp>" fallback for
* when it isn't loaded against a live deployment.
*/
const app = createApp()
const server = serve({ fetch: app.fetch, port: 0 }, async (info) => {
const base = `http://127.0.0.1:${info.port}`
const paths = {
health: '/v1/health',
chain: '/v1/chain',
stocks: '/v1/stocks',
stockDetail: '/v1/stocks/TSLA',
coins: '/v1/coins?limit=8',
launches: '/v1/launches?lookback=24h&limit=8',
portfolioChallenge: '/v1/portfolio/0xaF3D76f1834A1d425780943C99Ea8A608f8a93f9',
}
const samples: Record<string, unknown> = { capturedAt: new Date().toISOString() }
for (const [key, path] of Object.entries(paths)) {
const res = await fetch(base + path)
samples[key] = { status: res.status, body: await res.json() }
console.log(`captured ${key}: HTTP ${res.status}`)
}
// A <script src> file, not JSON fetched via XHR/fetch — the docs page must work when
// opened directly via file://, where fetch() of a local file is blocked by browser CORS
// policy but a <script> tag loads fine.
const outPath = fileURLToPath(new URL('../docs/samples.js', import.meta.url))
writeFileSync(outPath, `window.HOOD_API_SAMPLES = ${JSON.stringify(samples, null, 2)};\n`)
console.log(`wrote ${outPath}`)
server.close()
})