Typed SDK for the Builderforce.ai LLM gateway — chat completions with tool-calling and structured output, embeddings, image generation, models, and usage analytics over an OpenAI-compatible surface.
BuilderForce Agent Memory — engine layer. TypeScript/WGSL Mamba SSM kernels, model blocks, autograd, training, and BPE tokenizer (zero runtime deps, WebGPU-native).
BuilderForce Agent Memory — runtime layer. SSM execution, Transformer orchestration, online distillation, hybrid RAG retrieval, and persistent agent memory for BuilderForce.ai agents.
BuilderForce Agent Memory — MCP transport. Exposes @seanhogg/builderforce-memory to any MCP client (Claude Agent SDK in-process, stdio, or HTTP) over a pluggable MemoryBackend, with token-saving fetch-on-demand recall.
Headless client-side AI video-generation engine for the browser. Runs LCM / SD-Turbo diffusion on WebGPU or WebNN (onnxruntime-web), CLIP tokenization via @huggingface/transformers, Mamba SSM temporal coherence, and WebCodecs MP4 muxing. No React. For a r
Voice-cloning + LLM-narration client for Builderforce.ai Studio. One resolveNarrationEngine() seam routes text→speech through an on-device SSM clone engine (WebGPU, $0) or the metered gateway synthesize endpoint, with an honest named-voice fallback. Consu
Full embeddable AI Video Studio for React — drop-in that runs LCM / SD-Turbo diffusion on WebGPU/WebNN, conditioned on a Mamba SSM state for temporal coherence, and muxes MP4 via WebCodecs. React layer on top of the headless @seanhogg/builde