潜力新星（90天）

Whisper Studio is an independent desktop interface for OpenAI Whisper.

Convert long AI conversations into portable conversation state graphs for LLM handoffs.

Private, on-device AI desktop app — GGUF (llama.cpp) & MLX, with Qwen3.8's native reasoning-effort ladder wired in as a real control. Local coding agent, RAG knowledge base, Deep Research, vision and voice. 100% offline, no account, no telemetry. Windows & macOS.

RAMen is a fast in-memory data store like Redis, but built for AI: drop-in Redis protocol, native vector search, semantic caching, and a built-in MCP server for agents. Single Go binary, BSD-3.

A deterministic, zero-dependency prompt-pruning layer for long-running LLM conversations — expires stale tool state, collapses duplicate context, and proves it never drops a fact a later turn still depends on.

🎬 Director-grade Seedance 2.0 prompt OS for OpenClaw + Hermes agents — 28 skills, T2V/I2V/V2V/FLF2V, multi-clip sequences, zh/ja/ko support, IP-safe rewrites & pro delivery. Powered by MyClaw.ai

A continuously-verified dataset of free-tier & trial-credit LLM APIs for developers. Every entry dated, sourced to the provider's own docs, and machine-readable. No hype, no dead links.

Plug-and-play, dependency-free web command center for DGX Spark fleets (and any GPU boxes): live per-node GPU/CPU/mem/temps/power, optional RouterOS switch monitoring, and per-model decode/prefill/TTFT/KV cards. Point it at your own hardware.

GPU 性能与 AI Infra 学习项目：CUDA/Triton 算子、NCU/NSYS、vLLM/SGLang/TRT-LLM/ms-swift、PyTorch/DeepSpeed/ms-swift 训练、并行架构

SitRep — free, open-source, offline voice dictation. Press a shortcut, speak, and your words appear in any text field. Powered by Whisper/Parakeet, built with Tauri.

Real-time bilingual subtitles for any browser video. Offline-first speech recognition (Whisper + SenseVoice, 100+ languages) with local Ollama AI translation — 100% private, ultra-low latency.

AI agents now operate with authority. Authority without discipline is how complex systems fail. Nuclear’s control loop, ported to AI-assisted software engineering.

AI Paper Reviewer: LLM-based tool for academic paper review

AI academic PPT workflow skill for Codex / ChatGPT: paper-to-editable PowerPoint, planning table, mockup family, and PPTX generation.

Chain two LLMs — local or cloud: a Prompter that refines your idea, and a Coder that writes the code.

Make any Claude model adopt Claude Fable 5's working style in Claude Code — always-on output style + zero-dependency MCP + subagent injection. A style transplant, not a capability transplant.

Serve GLM-5.2 469B (REAP-pruned, NVFP4) across 3× NVIDIA DGX Spark with vLLM pipeline parallelism — 256K context, production-ready config and patches

Local LLM chat panel for ComfyUI — LM Studio & Ollama, multi-session, vision support, Guide Materials

A searchable media gallery and metadata viewer for ComfyUI.

OpenCode agent configuration — a read-only orchestrator delegates to 6 specialized subagents. Expert model plans, fast models execute. Isolated contexts, minimal permissions, adversarial verification built in. Ready to drop in.