플러그인

DSH의 다중 모델 접근, 월별 크레딧, 수동·품질·크레딧·자동 라우팅, 실패 시 재라우팅, 사용량 감사를 관리합니다.

dsh용 보안 감사 및 세션 포렌식입니다. 전체 도구 호출 기록(누가/어떤 도구/파일/시각), 민감정보를 가린 재생, 규칙 기반 스캔을 제공합니다.

세션 관리를 위한 DSH Web 플러그인입니다. 보관/보관 해제, 확실한 물리적 삭제(transcript 파일과 ~/storages/*.json 전체 검사), 드래그 앤 드롭 workspace 재할당, 브랜치 기반 메시지 편집(edit/retry/reroll 및 되돌릴 수 있는 버전 타임라인).

DeepSeek Harness를 위한 로컬 우선, 사람 승인 기반의 작가적 영상 워크플로

지원되는 파일 변경에 대한 턴 단위 설정 가능 예산: 도구 실행 전에 대상 파일 수, 변경 호출 수, UTF-8 페이로드 바이트를 제한합니다.

모든 DSH 라우트에 비전을: 웹 컴포저에 이미지를 붙여넣으면 의도를 반영해 자동 분석하고, 워크스페이스 이미지 읽기는 Kimi/MiniMax 비전 서브에이전트에 위임하며, 붙여넣은 원본은 편집용으로 실체화합니다.

웹 UI용 Google AI Studio 스타일 플로팅 미리보기: 화면비와 해상도 프리셋이 있는 드래그/크기 조절 가능 iframe 미리보기, 그리고 직접 편집, 확인 후 쓰기, 세션 로그 기록을 지원하는 VS Code 스타일 코드 창.

현재 dsh 세션의 실시간 명령 체인과 프로젝트 스킬을 확인·관리, 그 자리에서 생성 및 편집 가능 — 보이는 그대로가 모델에 전달됨

DeepSeek Harness용 자가 편집 데스크톱 워크벤치: dsh web을 완전히 재현한 편집 모드에서 에이전트가 핫 리로드로 워크벤치(패치 레이어, 플러그인, UI)를 직접 수정

Monitors Windows console windows and hides or closes empty ones — consoles whose interpreter process has no business child process left besides conhost or openconsole and stays idle at about zero CPU. Recognizes both classic console windows and Windows Terminal CASCADIA tabs that only host a bare cmd, powershell, pwsh or python prompt. Hard-protects Fluent, MPI and solver processes, the DSH host, Explorer and the foreground window, so running solvers are never touched. Also provides a taskbar window cleanup action named close other windows, plus a persistent exemption whitelist of process names and title keywords stored in ~/.dsh/console-sentinel/whitelist.json. A settings panel under Settings then Console Sentinel shows live window status, one click hide of empty consoles, a taskbar cleanup preview and run button, and whitelist editing. Windows only.

Image Editor separates uploaded images into editable layers through Shenhuo MCP, with confirmed credit quotes, browser editing and local PSD downloads.

Adds seven `art_*` agent tools plus a right-sidebar drawing workbench (series, pages, revisions, regions, camera blockout) that submits ComfyUI graphs to a ComfyUI instance you host, for MOSS/K2 text-to-image, inpaint, character references, region editing and Qwen-Image edits, with up to four caller LoRAs stacked per request. Requires a local checkout and one adapter step: after a marketplace install, run `install/adapt.mjs` (or use `install/install.ps1` on Windows) to replace the `__MOSS_*__` placeholders with an absolute `workspaceRoot` and a ComfyUI origin; until then the Host refuses to load the preset with `MOSS_CONFIG_*`.

One-step local model onboarding plus a light harness for DSH. On load it probes loopback and private-network inference endpoints (llama.cpp, Ollama, vLLM, LM Studio), reads each model's context window and output budget from the backend, and registers the provider through settings, so a local model becomes selectable without hand-writing config. It then trims only that provider: masks the tool list through tools.restrict, drops named system-prompt sections and memory contexts, swaps the injected instruction block for a short one while preserving its first 80 and last 96 characters, and truncates tool-schema descriptions without touching structure or required fields. Cloud providers are returned as the same assembly object, untouched. The injected block is re-read from disk on every assembly, so editing the file takes effect on the next request with no reload. The panel in the plugin detail page is a live operations view: it shows whether the takeover is active and by which hook, how many characters the last assembly was trimmed from and to, and which providers are online (probed concurrently, with a real one-shot completion reporting latency and reply text per provider). If a local backend is not running, the panel offers a one-click launch: it probes the port first (never starting a duplicate), then starts the matching server (ollama serve, or a custom command you supply) detached and polls until the endpoint answers, clamping the wait to 3-600 seconds, reaping the child it just spawned if it never answers, and naming the paths it searched when it cannot find the binary. A stop button kills backends this plugin started (SIGTERM, then SIGKILL) and refuses to touch a server somebody else launched. When a model has just been downloaded and no backend exists yet, the panel can also scan the machine for .gguf files with hard caps (300 models, 20000 directory entries, Downloads only one level deep), start llama-server on the chosen one (skipping the .mmproj. vision projection and reusing a server that already answers), register the provider under the model id the backend reports, and run one real completion to prove it works. Before spawning anything it does a fit check that counts only the part of the file the GPU cannot hold: hostBytes = fileSize - VRAM, compared against max(total RAM - 2GB, free RAM). A file far larger than RAM is refused outright; a heavy host-side remainder is refused unless you confirm; the same remainder under mmap is only a warning, because the weights land in reclaimable page cache. A load-mode selector (auto / mmap / none / mlock / mmap+mlock) maps to llama.cpp --load-mode, and the default is whatever the engine picks. From 0.4.2 everything above is derived from the platform instead of assumed: home and data directories come from os.homedir() and the OS conventions (LOCALAPPDATA on Windows, ~/Library/Application Support on macOS, XDG_DATA_HOME elsewhere); backend binaries are looked up in the official install locations of each OS and can be pinned with OLLAMA_BIN, LLAMA_SERVER_BIN, VLLM_BIN or LMSTUDIO_BIN; model discovery walks DSH_LOCAL_MODELS_DIR, ~/models, ~/llama.cpp/models, ~/.cache/llama.cpp and the LM Studio data directory, with the WorkBuddy layout demoted to one candidate among them. VRAM is asked from nvidia-smi, then rocm-smi, then sysctl hw.memsize on Apple Silicon (flagged unified, so the same bytes are not charged twice), and any flip to a unified-memory budget is handled explicitly. Before building a command line the plugin runs llama-server --help once and keeps only the flags that build actually has, so older engines get --n-gpu-layers 99 and no KV quantization while newer ones get auto plus q8_0; anything it cannot probe is passed through rather than dropped. LM Studio is started with lms server start --port N rather than being lumped in with the engines that need a model argument, and vLLM gets its own --model example instead of llama.cpp syntax. Windows, macOS and Linux are all exercised in CI, including a windows-latest and a macos-latest job, because a cross-platform claim that has only ever run on one OS is not a claim. Two of those suites exist specifically to catch this machine's blind spots: one replays the whole suite with process.platform rewritten to each OS, and one deletes every advantage this development box has — no nvidia-smi, no rocm-smi, no sysctl, an empty PATH, a home directory with no models — and requires graceful degradation rather than a crash. That last one caught a real class of bug: several assertions hardcoded 'lms.exe', which cannot pass where the executable has no suffix. Measured on the reference machine: a 16.9GB 27.2B MoE with 14.5GB of VRAM holds a 285-691MB working set and sustains 76 tok/s after a 1000-2000 token warm-up, so file size alone is not a fit signal. From 0.4.3 the scanned models are shown as a table rather than a dropdown (one row per gguf, with its size and any fit warning), so you can see what is on disk before choosing. From 0.4.4 the delete button on the provider table removes the provider entry, not the weights file on disk. That distinction matters: what people actually want to clear out is the dead entry - one pointing at a port that never answered, one registered twice for the same port, one whose model file was deleted long ago - not a 15.8GB gguf. Deleting a config is undone by importing again; deleting a gguf is not. The removal goes through settings.mutate with an unset op on \['providers', id], because update merges layers: writing null there leaves a null provider key that the loader reads as a broken config rather than an absent entry, and only mutate/unset is implemented as Reflect.deleteProperty. The result is read back afterwards, because unset on a key that comes from a lower layer (the profile cordis.patch.yml or a home patch) does not delete it - it restores the inherited value, so the call succeeds while the entry stays. When the entry is still there the panel says so and points at the file to edit instead of reporting success, and a DSH whose settings service predates mutate() is told plainly rather than silently ignored. Deleting takes two clicks and clearing the confirmation is as easy as clicking another row; on success the online and baseURL caches are cleared and the tool surface is recomputed, since the provider list changed. The gguf itself is never touched by this plugin - the scanned-model table is display and selection only.VRAM reporting is driver-dependent. From 0.4.5 model discovery also reads Ollama's own storage: its blobs are named sha256-… with no extension, so extension matching could never list them. Weights are now accepted by extension OR by GGUF magic header (only dot-free files of at least 8MB, capped at 600 header reads per scan; obviously-not-weights extensions like .json/.md/.safetensors/.part are never opened). Ollama models are resolved from their manifests: the model layer is picked by mediaType application/vnd.ollama.image.model (never adapter/mmproj layers), the human-readable name is restored from the manifest path (…/qwen3/8b becomes qwen3:8b, …/latest stays bare), tags sharing one blob merge into a single row, and the metadata kv supplies name, quantization and context length. Adopting an Ollama model does not start a duplicate llama-server over the same weights: it confirms ollama serve answers on 11434 (launching it if not) and registers the provider under the tag the endpoint itself reports — the blob's sha256 name would never match anything, since 11434 lists qwen3:14b, not sha256-a8cc…. Two real-machine-only traps fixed on the way: existsSync('<file>/') returns true on Windows, so manifest-vs-directory branching now goes by parseable content instead of trailing-slash semantics; and the quantization label table had been copied from a stale ggml enum generation, labelling ggml_type 15 as Q5_K_S while ollama show reports Q4_K_M. Same-blob dedup is by (dev,ino), keeping whichever entry has a name. Verified on the reference machine: the scan goes from 1 row to 4 — qwen3:14b (8.64G) and qwen3:8b (4.87G) now appear with Q4_K_M badges — in 4ms; npm run verify exits 0 (launcher 57, portable 35, panel-host green including the linux/darwin simulated-platform runs); the new panel-host suite builds a real miniature Ollama store in tmpdir and points OLLAMA_MODELS at it, so results no longer depend on whether the test machine happens to have Ollama installed. From 0.4.6 the panel shows each model its real name instead of its file name, because some download tooling stores every weight under one generic file name, leaving the real name only inside the GGUF header. The scan now reads each file GGUF KV section - not a fixed-size head buffer, because tokenizer.ggml.tokens (hundreds of thousands of strings on a large MoE model) and the chat template push general.file_type past any window; and GGUF string arrays are interleaved (u64 length plus payload per element), so treating them as a length table walks the cursor off the rails and produces astronomically large reads. Reading and skipping are accounted separately: key names and types are always read, string values over 256KB and string arrays are only walked past without copying, non-string arrays over 64 elements are skipped outright, and a cursor past the file size stops the read, because an out-of-range position means the cursor is wrong rather than that a limit should be raised. The displayed name goes Ollama tag, then general.basename plus size_label, then general.name, then the file name, with the full string kept for the hover text since it usually carries internal build numbers; mmproj files are now identified by general.type as well as by name. A short read (truncated file) no longer feeds half a value into the parser. On the reference machine, reading one file header parses 44 keys in about 16ms, and a whole scan takes about 31ms. From 0.4.7 the smoke test no longer fails the first time and passes the second. The old code gave both request attempts one shared AbortController and one shared 30-second timer, so the total budget covered both rounds: a cold load of a large model burned it in round one, the abort left the signal permanently aborted, and the second round (the Ollama-style reasoning_effort: none retry) could not even be sent - which is exactly why the first click returned This operation was aborted and the second succeeded. Now each attempt gets its own controller and timer, a timeout retries once automatically (retryOnTimeout, on by default) because a first-request timeout almost always means the weights are still being read from disk, and non-timeout errors (refused connection, HTTP error) are not retried at all. Timeouts report themselves: the message says the wait exceeded smokeTimeout, whether a retry already happened, and points at the new smokeTimeout setting (seconds, default 60, clamped to 5-600) - port responding is not the same as tokens coming out, so it is a separate knob from launchTimeout. Unrelated: an example naming one specific model was dropped from the panel copy, since not every machine has that model downloaded, and the explanation now describes the mechanism instead.

WYSIWYG Markdown notes editor in the DSH right sidebar, with tabbed documents, table editing, chat integration, assistant read/write tools, and local file export.

A floating sticky-note pad for the DeepSeek Harness Web GUI with a note directory, text editing, pasted images and local persistence.

Prompt manager for DSH: imports prompts from JSON or Markdown sources and DSH's own system sections, then controls what is injected into the system prompt — per prompt, or per agent preset — with title editing, drag ordering, LLM translation and the option to inject either the original text or its translation.

DeepSeek Harness Agent preset 관리자입니다. DSH 설정 파일을 편집하지 않고 실행 중 agent preset을 등록하고 관리합니다.

로컬 영상 색상 편집: 카메라에 맞춘 DJI Log LUT, 독립 하이라이트/그림자, 선택형 피부/선택 색상 제어, 프레임 미리보기, 10비트 HEVC 내보내기를 제공합니다. 여기에 등록된 universal tarball에는 runtime이 없습니다. 최초 “configure environment” 버튼은 고정 플랫폼 runtime(Python 3.12 + FFmpeg 9, 약 100MB)을 HTTPS로 받거나 자체 Python 3.10+ 및 FFmpeg/ffprobe 8+를 지정합니다. 플랫폼별 release 패키지에는 runtime이 포함됩니다.

전송한 메시지를 우클릭해 편집하거나 철회합니다. 실행 중 답변을 먼저 중지한 뒤 해당 메시지와 이후 내용을 대화에서 교체합니다. 철회는 짧은 철회 표시를 남기고 거기서 멈추며, 편집은 새 텍스트로 모델에 다시 답하도록 요청합니다.

자주 쓰는 한 줄 문장을 composer 위 캡슐에 저장합니다. 사용 빈도에 따라 재정렬되며 {{placeholder}} 값을 채운 뒤 초안에 삽입합니다. 클릭은 교체가 아니라 추가라서 작성 중인 초안이 유지됩니다. 서랍에서 검색/그룹화/편집하고 수동 편집 가능한 Host 측 JSON에 저장합니다.

DeepSeek Harness용 로컬 작업대입니다. model 지표, 비용 추정, prompt 프리셋, 선택적 요청 감사 기능을 제공합니다.

Codex 호환 Skill Studio입니다. 왼쪽 Sidebar에서 SKILL.md skill을 탐색/생성/편집/활성화/비활성화/실행하고 `~/.codex/skills` 및 프로젝트 `.codex/skills`를 DSH에 연결하며 채팅의 `$skill-name` 호출을 지원합니다.

오른쪽 Sidebar 문서 미리보기에서 텍스트를 바로 편집하고 ⌘S로 저장하며 외부 변경 충돌을 확인합니다. Workspace Files 탭은 퍼지 검색, 생성/이름 변경/삭제, 파일별 Git 상태를 지원합니다. Git changes 탭은 remote upstream을 기준으로 하며 원격과 일치하는 파일은 목록에서 빠집니다.

자체 Lua script API로 로컬 Synthesizer V Studio 2 Pro를 에이전트에 연결합니다. 검색, 상시 브리지, 보컬 구성, 음표 편집, DAW(VST3/ARA + MIDI) 연동 도구 5개를 제공합니다. Windows 10/11 전용이며 Synthesizer V Studio 2 Pro 2.1.2 이상과 열린 프로젝트가 필요합니다.