Plugins

Realtime WebGPU ocean skin with quick controls for waves, daylight, glass opacity, and an automatic day cycle.

Research-assistant plugin suite (literature search, research wiki, LaTeX compile, independent subagent review) for the DeepSeek Harness

ROS2 debugging toolset and robot-state vision analysis for DeepSeek Harness: node/topic/service/action/interface/TF enumeration, whole-graph topology JSON, rosdep checks, approval-gated builds and custom message scaffolding, GUI screenshots and multimodal vision observation, plus headless RViz2 offscreen rendering (low-poly meshes, direct pixel read, GPU passthrough - motion rendering at 30Hz) with parallel VLM realtime analysis.

Large-image recognition for vision-exp models: lossless 800×800 tile recognition (smart/pipeline/full), local OCR with preprocessing & handwriting routing, optional multi-vendor GPU (DirectML/CUDA/OpenVINO) with auto CPU fallback.

Renders Wallpaper Engine scene wallpapers in real time in the browser (three.js player plus a Rust/WASM CPU particle simulation) and plays video and web wallpapers, falling back to the preview image, with no Wallpaper Engine runtime required. Scans the local Steam workshop library (431960) by auto-detection or a manual path, and offers a settings panel for wallpaper selection, library and engine directories, glow, pause and quality.

Fourth permission preset for dsh: unconfined, GPU-capable sessions (danger-full-access) with per-operation user approval for writes outside the workspace or to protected paths (.git/**, .env*); implemented purely as a bundle over the official tools/pre-execute ask hook, no core edits.

Purple-white desktop pet that docks to the screen edge and shows real-time CPU, RAM, GPU and NVMe temperatures with hardware info.

A dockable IDE workbench for the DSH web UI: a GPU-accelerated terminal (restty renderer over a Rust PTY), a CodeMirror editor, a file explorer, notes, a sandboxed browser, and ripgrep content search.

Sub-agent matrix swarm that routes heterogeneous tasks to the most suitable model from an OpenRouter-like gateway plus cfgpu.com/llm/square, dispatches each via in-process subagents or direct LLM calls, and tracks per-model token consumption with a feedback-driven ranking.

DeepSeek Harness (dsh) plugin: live NVIDIA GPU monitor in the web rightbar. NVML-first on Linux (nvidia-smi fallback). NVIDIA only! Not AMD/Intel/macOS. Multi-GPU util/VRAM/power/temp, processes, sparklines. Installable dsh.bundle.

Reports Docker CLI, context, and daemon reachability in WSL, plus vLLM :8000 /v1/models health and GPU runtime hints.

Probes nvidia-smi and GPU visibility inside WSL.

GPU-aware execution layer for DeepSeek Harness: gpu_status / gpu_exec / gpu_run_bg tools, per-step GPU context injection, automatic CUDA_VISIBLE_DEVICES card selection

Read-only Windows device information tools: one agent tool per Win32 device category (time, system, cpu, memory, disk, gpu, network, battery, processes, usb, audio, printers) via WMI/CIM and Node os collectors.

Live btop-style CPU/memory/temperature/GPU monitor tab for the DeepSeek Harness web GUI.

Multi-instance ComfyUI orchestration: probes every configured ComfyUI endpoint for its GPU, free VRAM and queue depth, then dispatches a generation job to the idlest eligible instance. Ships three tools (comfyui_farm_status, comfyui_farm_pick, comfyui_farm_run). An instance is eligible only if it is reachable, has at least the requested free VRAM, and its queue is no deeper than the configured ceiling; eligible instances are ranked by free VRAM minus queue depth times a weight, so an idle shallow queue wins over a busy large one. Zero runtime dependencies. Optional retry on probe failure, because a single timeout across an SSH tunnel is not proof that an instance is down.

AlphaRSI — three-role self-improvement loop that evolves the training recipe of alpha sequence models. Operator runs walk-forward, Evaluator grades with OOS RankIC, Refiner proposes hyperparameters via TPE Bayesian search; only strictly-better recipes are kept. Offline one-command run, data bundle shipped with sha1 reconciliation, GPU auto-detect. Ships one skill.

Large-image recognition for vision-exp models: lossless 800×800 tile recognition (smart/pipeline/full), local OCR with preprocessing & handwriting routing, optional multi-vendor GPU (DirectML/CUDA/OpenVINO) with auto CPU fallback.

Adds /dev/dxg and /proc read-write grants to the DSH Landlock sandbox on WSL2, so CUDA is available inside a confined agent bash session.

A switchable WebGPU black-hole theme for the DeepSeek Harness Web UI that retains the built-in Dark palette and animates only the new-session background.

Floating GPU monitor for the DSH Web UI: a compact, draggable strip with one row per discrete GPU (exact name, utilization, VRAM, temperature, power draw) from nvidia-smi, stays visible over side panels and side cards, with a top-VRAM-process list when the driver reports it; degrades to an n/a pill on machines without an NVIDIA driver.

Keeps a local model's context at a size the GPU handles well: measures prefill speed on every request and, before each step, warns or compacts early on a hard token ceiling, an observed time-to-first-token budget, or a predicted cold-prefill budget; /context-budget shows the numbers and the cost of compacting now.

Control panel for local LLM inference servers in the DSH web view: service registry with health checks and stop controls, named parameter profiles for llama.cpp, SGLang and vLLM with pre-save KV/VRAM validation, GPU detection, and tok/s benchmarks.

DSH plugin (dsh-desktop-launcher) that launches the dsh-desktop Windows Electron shell from the conversation: double-click to run the DSH web in a standalone window, region screenshot with GPU-accelerated selection that auto-pastes into the chat (works with vision bridges), system tray, session manager, auto environment setup, configurable port, single-instance lock.