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
// GET /api/cron/mirror-fanout - drive the custodial copy-trade (mirror) engine.
//
// For every active follow edge whose leader has made a NEW confirmed trade since
// the edge's cursor, size the trade for the follower and execute it through the
// task-05 engine inside the follower's spend policy (api/_lib/agent-mirror.js).
// Idempotent end to end: the agent_mirror_fills unique key and the custody
// idempotency key both prevent double-mirroring, so re-running this cron (or
// overlapping with an owner "Sync now") never double-spends.
//
// Bounded so a 2-minute cron can never run away: only edges with a leader trade
// in the recent window are scanned, and each edge processes at most N events.
import { json, method, wrapCron } from '../_lib/http.js';
import { sql } from '../_lib/db.js';
import { syncFollow } from '../_lib/agent-mirror.js';
import { requireCron } from '../_lib/cron-auth.js';
const NETWORKS = ['mainnet', 'devnet'];
const MAX_FOLLOWS_PER_RUN = 120;
const MAX_EVENTS_PER_FOLLOW = 15;
async function fanout(network, stats) {
// Candidate edges: enabled, not killed by the follower's agent-wide switch, and
// whose leader has at least one confirmed trade newer than the edge cursor.
const follows = await sql`
SELECT f.*, la.name AS leader_name
FROM agent_mirror_follows f
JOIN agent_identities fa ON fa.id = f.follower_agent_id AND fa.deleted_at IS NULL
JOIN agent_identities la ON la.id = f.leader_agent_id AND la.deleted_at IS NULL
WHERE f.enabled = true
AND f.network = ${network}
AND COALESCE(fa.meta->>'mirror_killed', 'false') <> 'true'
AND EXISTS (
SELECT 1 FROM agent_custody_events e
WHERE e.agent_id = f.leader_agent_id
AND e.network = ${network}
AND e.category = 'trade'
AND e.status = 'confirmed'
AND e.id > f.last_leader_event_id
AND e.created_at > now() - interval '20 minutes'
)
ORDER BY f.updated_at ASC
LIMIT ${MAX_FOLLOWS_PER_RUN}
`;
stats.edges = (stats.edges || 0) + follows.length;
if (!follows.length) return;
for (const f of follows) {
try {
const r = await syncFollow(f, { maxEvents: MAX_EVENTS_PER_FOLLOW });
for (const res of r.results) {
stats[res.status] = (stats[res.status] || 0) + 1;
}
} catch (err) {
stats.error = (stats.error || 0) + 1;
stats.last_error = (err?.message || 'error').slice(0, 160);
}
}
}
export default wrapCron(async (req, res) => {
if (!method(req, res, ['GET', 'POST'])) return;
if (!requireCron(req, res)) return;
// Seeded so an idle run answers {edges:0} rather than a bare {ok:true}: a
// scan that found nothing has to be distinguishable from a scan that never
// reached the query.
const stats = { edges: 0 };
for (const network of NETWORKS) {
try { await fanout(network, stats); }
catch (err) { stats[`error_${network}`] = (err?.message || 'error').slice(0, 160); }
}
return json(res, 200, { ok: true, ...stats });
});