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
// Local pose resolution: the same deterministic algorithm the hosted
// pose_model tool runs behind POST /api/mcp-3d, executed in-process over the
// bundled preset library. Ported verbatim from the server implementation
// (api/_mcp3d/tools/studio.js in the three.ws repo) so a zero-config
// poseSeed() needs no network, no key, and no payment, and resolves
// byte-identically to the hosted tool: same token scoring, same sha256
// fallback pick, same seed derivation.
//
// Hashing uses WebCrypto (crypto.subtle), available in Node 18+ and every
// modern browser, so the package stays dependency-free on both runtimes.
import { PRESETS, PRESET_GROUPS, getPresetById } from './pose-presets.js';
function tokensOf(str) {
return String(str || '')
.toLowerCase()
.split(/[^a-z0-9]+/g)
.filter(Boolean);
}
const INDEX = PRESETS.map((preset) => {
const idTokens = tokensOf(preset.id);
const labelTokens = tokensOf(preset.label);
const groupTokens = tokensOf(preset.group);
return {
preset,
all: new Set([...idTokens, ...labelTokens, ...groupTokens]),
idTokens,
labelTokens,
};
});
function scorePreset(promptTokens, entry) {
let score = 0;
for (const t of promptTokens) {
if (entry.all.has(t)) score += 3;
else {
for (const tok of [...entry.idTokens, ...entry.labelTokens]) {
if (tok.includes(t) || t.includes(tok)) {
score += 1;
break;
}
}
}
}
return score;
}
async function sha256Bytes(text) {
const digest = await crypto.subtle.digest('SHA-256', new TextEncoder().encode(text));
return new Uint8Array(digest);
}
function toHex(bytes) {
let out = '';
for (const b of bytes) out += b.toString(16).padStart(2, '0');
return out;
}
async function pickPreset(prompt) {
const tokens = tokensOf(prompt);
const deterministic = async () => {
const hash = await sha256Bytes(String(prompt));
// Same pick as the server's hash.readUInt32BE(0): big-endian first word.
const idx = new DataView(hash.buffer, hash.byteOffset, hash.byteLength).getUint32(0) % INDEX.length;
return { entry: INDEX[idx], score: 0, reason: 'no-match-deterministic-pick' };
};
if (tokens.length === 0) return deterministic();
let best = null;
let bestScore = -1;
for (const entry of INDEX) {
const sc = scorePreset(tokens, entry);
if (sc > bestScore) {
best = entry;
bestScore = sc;
}
}
if (bestScore <= 0) return deterministic();
return { entry: best, score: bestScore, reason: 'token-match' };
}
/**
* Resolve a prompt to the tool's snake_case structuredContent shape, so the
* local and remote lanes normalize through the exact same `shape()` step.
* @param {string} prompt
* @returns {Promise<object>} structuredContent (snake_case, as on the wire)
*/
export async function resolvePoseLocal(prompt) {
const picked = await pickPreset(prompt);
return shapeStructured(prompt, picked.entry.preset, { score: picked.score, reason: picked.reason });
}
/**
* Resolve a preset DIRECTLY by id, bypassing prompt scoring. The scoring
* pipeline can shadow an id (group vocabulary ties, first preset wins), which
* would violate presetPose()'s "resolve this exact preset" contract; a direct
* lookup cannot. The seed uses the same formula with the id as the prompt, so
* it matches a scoring run that picks the same preset.
* @param {string} presetId
* @returns {Promise<object|null>} structuredContent, or null for an unknown id
*/
export async function resolvePresetLocal(presetId) {
const preset = getPresetById(presetId);
if (!preset) return null;
return shapeStructured(presetId, preset, { score: 0, reason: 'preset-id' });
}
async function shapeStructured(prompt, preset, match) {
const seed = toHex(await sha256Bytes(`${prompt}|${preset.id}`)).slice(0, 16);
return {
seed,
preset_id: preset.id,
preset_label: preset.label,
group: preset.group,
parameters: preset.pose,
preview_url: null,
match,
groups: PRESET_GROUPS,
};
}