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
docs(crews): document the WebGL context budget and the still-portrait fallback
nirholas5e0924c506cb
diff --git a/docs/crews.md b/docs/crews.mdindex f709326e5..d7c26efdf 100644--- a/docs/crews.md+++ b/docs/crews.md@@ -42,6 +42,15 @@ in [`src/crews-page.js`](../src/crews-page.js) caps how many members render as alive `<agent-3d>` (6); everyone past that renders as their avatar's still image,and live figures mount only when they scroll into the stage.+Two rules keep those contexts alive once they exist. Presence refreshes every 20+seconds, and the room writes the new state into the figures already standing+rather than rebuilding the stage: the cast is rebuilt only when the membership+itself changes, so a long-open tab is not creating and discarding WebGL contexts+four times a minute. And a figure whose model fails to load (the agent's GLB was+replaced, storage had a bad minute) falls back to the member's still portrait on+the same plinth, so the person keeps standing in the room instead of leaving a+gap.+## APIAll endpoints answer `{ data: … }` on success and