Tag
Explore by tags
Composes AI agent teams from a Ghost+Shell+Model formula: each Bot pairs a prompt/MCP/Skills Ghost with a Chat, ClaudeCode, or Dify shell and a model like Claude or DeepSeek. Bots form Teams that run as traceable Tasks, wired to GitHub and DingTalk.
Extends vLLM beyond text to serve omni-modal models — Qwen3-Omni, TTS like CosyVoice3, and diffusion image/video/audio generators — in one engine, adding the non-autoregressive Diffusion Transformer support the core project never targeted.
Provides Gymnasium-style APIs and tooling to run isolated, networked execution environments for agentic reinforcement learning. Offers async/sync EnvClients, Docker/Kubernetes container providers, a web UI and CLI for scaffolding and deploying environments (Hugging Face Spaces); experimental and evolving.
Provides a Gymnasium-style API and tooling to create, deploy, and interact with isolated execution environments for agentic RL training. Includes async/sync clients, a web interface, CLI, Docker-based deployment, and Hugging Face Spaces integration.
Runs text-to-video, image-to-video, text-to-image, and image editing inference with acceleration, offloading, quantization, and distributed execution for large visual generation models.
Coordinates about a dozen role-based AI agents — analyst, architect, developer, QA, scrum master — through a CLI, taking a feature from PRD and architecture docs into an automated dev cycle. Runs inside Claude Code, Cursor, Codex, or Gemini.
Provides a CLI-first framework to orchestrate autonomous AI agents and development workflows. Includes role-based agents, the ADE execution pipeline, IDE hooks and an NPX installer for quick setup—best for teams automating planning→development→QA.
Orchestrates low-latency, multi-stage pipelines for omni and multimodal models by running each stage with its own scheduler and using zero-copy shared memory for tensor transfer. Emphasizes per-stage bottleneck tuning and OpenAI-compatible streaming endpoints, suitable for TTS and multimodal serving.
Enables research-grade character animation with neural networks in a single NumPy/PyTorch environment — train models, run inference, and visualize results without leaving Python. Includes ECS-style architecture, mocap import (GLB/FBX/BVH), built-in renderer, and headless/standalone modes for rapid prototyping.
Train robot reinforcement-learning agents with a heterogeneous runtime that streams CPU-parallel physics simulations (MuJoCo / Motrix) via shared memory into GPU/accelerator policy learners; provides a unified CLI, cross-platform backend support and demo checkpoints.
A TypeScript framework for building programmable, headless autonomous agents with a harness-centric runtime. Includes an SDK and CLI, virtual sandboxes (just-bash) with optional full container sandboxes, provider-agnostic model settings, and connectors for CI/Daytona/MCP—suited for deployable agent runtimes.
Framework for running agents inside real applications — it exposes shared actions, SQL-backed state, tools, skills, jobs and UI surfaces so agents can act on app state instead of just chatting. Backend-agnostic TypeScript stack with cloneable app templates and visual planning/recap features.