AI Runtime & Agent Substrate
generation-context
Token-budgeted context window manager — system prompt, RAG, user history, scratchpad. Picks what to drop when the window fills.
247,401 / 1,000,000 tokens· 24.7%
claude-sonnet-5system4%
history6%
rag-context3%
prompt1%
buffer1%
truncation via reciprocal rank
12ms
More in nebutra.com
- 3d-pipeline3D asset generation + render — text-to-mesh + USD export + Three.js preview, with GPU-pool routing on the backend.
- agent-runtimeLower-level agent execution — message reducer, tool invocation, streaming state machine. The kernel @nebutra/agents wraps.
- agentsMulti-step agent runtime — Vercel AI SDK foundation, tool registry, streaming UI, MCP integration. The skeleton for every Sailor agentic feature.
- ai-primitivesShared helpers used across all AI packages — token counting, prompt templating, stream-parser, retry-with-backoff. No provider dependency.
- ai-providersProvider metadata — pricing per million tokens, context windows, modalities. Drives model selection and cost estimates.
- atelier-canvasInfinite design canvas — multiplayer cursors, layer panel, frame export. The drawing surface every atelier play renders into.