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 { describe, it, expect } from 'vitest';
import { readFileSync, existsSync } from 'node:fs';
import { resolve } from 'node:path';
import { inspectGlb, glbJsonChunkEnd } from '../api/_lib/glb-inspect.js';
const avatar = (name) => resolve(process.cwd(), 'public/avatars', name);
// Real bundled GLBs: rigged humanoids + one static mesh.
const RIGGED = ['cesium-man.glb', 'fox.glb', 'realistic-female.glb'];
const STATIC = ['mannequin.glb'];
describe('inspectGlb allowPartial (ranged-read prefix)', () => {
for (const name of [...RIGGED, ...STATIC]) {
const path = avatar(name);
const present = existsSync(path);
it.runIf(present)(`prefix inspection matches full inspection — ${name}`, () => {
const full = readFileSync(path);
const fromFull = inspectGlb(full);
expect(fromFull).toBeTruthy();
// A prefix covering exactly the JSON chunk yields the same rig verdict,
// even though declaredLen (the full file size) exceeds the prefix length.
const end = glbJsonChunkEnd(full);
expect(end).toBeGreaterThan(20);
const prefix = full.subarray(0, end);
const fromPrefix = inspectGlb(prefix, { allowPartial: true });
expect(fromPrefix).toBeTruthy();
expect(fromPrefix.isRigged).toBe(fromFull.isRigged);
expect(fromPrefix.skeletonJointCount).toBe(fromFull.skeletonJointCount);
expect(fromFull.isRigged).toBe(RIGGED.includes(name));
});
}
it.runIf(existsSync(avatar('cesium-man.glb')))(
'a prefix too short to cover the JSON chunk returns null (refetch signal)',
() => {
const full = readFileSync(avatar('cesium-man.glb'));
const tiny = full.subarray(0, 16);
expect(inspectGlb(tiny, { allowPartial: true })).toBeNull();
// …and a small (≥20-byte) header prefix is enough for glbJsonChunkEnd to
// report how many bytes to refetch — it only reads the chunk-length field.
const headerOnly = full.subarray(0, 64);
expect(glbJsonChunkEnd(headerOnly)).toBe(glbJsonChunkEnd(full));
},
);
it('full-buffer mode still rejects a truncated buffer (declaredLen > length)', () => {
const path = avatar('fox.glb');
if (!existsSync(path)) return;
const full = readFileSync(path);
const truncated = full.subarray(0, glbJsonChunkEnd(full)); // missing BIN chunk
// Without allowPartial the declaredLen consistency check rejects it.
expect(inspectGlb(truncated)).toBeNull();
// With allowPartial it parses fine.
expect(inspectGlb(truncated, { allowPartial: true })).toBeTruthy();
});
});