Plugins

Replaces the sidebar workspace Files pane with a file manager that adds real composer chips for "@" file mentions, word and line selection across syntax highlighting, archives in and out, upload and download, and a jump menu to any directory on the machine.

Left-sidebar panel that browses the filesystem of the machine the Harness runs on: directory listings with sizes and modification times, text and image previews, file downloads, file uploads into the open directory, and folder creation. The panel speaks twelve languages. Reads and writes run with the privileges of the account that runs the Harness.

本插件仅供娱乐使用，非实用性插件，下载前请自行选择。DeepSeek Harness 非官方社区插件（Token-Reducer?）：模拟执行任务，生成 SIMULATED 轨迹与 0 字节产物，不调用真实模型或工具/For entertainment only. Not a practical plugin — download at your own discretion.Unofficial DeepSeek Harness plugin that simulates task execution with SIMULATED traces and zero-byte artifacts.

Feishu/Lark integration for DeepSeek Harness through the official lark-cli: a sidebar panel for app setup and QR sign-in, a command console, and agent tools that run any lark-cli command across 18 business domains, 200+ high-level commands and the full OpenAPI; the official CLI is downloaded on first use on Windows, macOS or Linux.

Video workbench for DeepSeek Harness that parses a link, downloads the source, transcribes it, trims or merges clips and exports GIFs. Requires ffmpeg/ffprobe and yt-dlp; whisper.cpp plus a model is optional for videos without subtitles.

A local ComfyUI workbench for DeepSeek Harness (DSH): image/audio/video generation, W2/W3 sprite cards with Qwen back-view walk-up, local Flux.2 Item Studio assets with guided SDXL fallback, Hugging Face downloads, inspection and arbitrary workflows.

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

Windows multi-drive dashboard that labels cleanable, caution, and system space in green, yellow, and red, audits download folders, and suggests how to split data across drives.

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. From 0.4.8 adopting a plain .gguf no longer registers an unrelated Ollama model. The port field used to follow the last click rather than the file: selecting an Ollama entry moved it to 11434, selecting a plain .gguf did not move it back, so a 11434 left over from an earlier click stayed. The server accepted it as given, probed the port, found Ollama answering and borrowed it (the skipped-launch log line about the port already answering is this path), and then the model id - derived from the file name - could not be matched against Ollama's /v1/models, where the fallback was to grab the first entry in the list. Embedding models tend to sort first, so the adopt registered a model the user never selected. Now the port follows the file: a plain .gguf landing on another engine's port (ollama 11434, vLLM 8000, LM Studio 1234) is moved to 8090 and the panel says so. Only ports belonging to a known engine are overridden - a hand-typed 9999 aimed at your own llama-server is left alone. The model-id fallback no longer guesses either: it takes the only entry when there is exactly one (a single-model backend may legitimately report a different id than the file name) and otherwise reports an error listing the ids it did see, rather than registering a wrong provider.

Turn a logged-in DeepSeek web account into an OpenAI and Anthropic compatible endpoint, with tool calling, image input, an account pool and a local admin console. First use needs a one-time `node scripts/fetch-assets.mjs` in the installed plugin directory to download DeepSeek's PoW runtime from the vendor CDN; without it the proxy starts but every upstream request fails.

Adds a Share to chat action under each reply: converts the answer to WeChat or WeCom text, ordered segments or a long image, editable-table email HTML, or a downloadable Word document.

Instala, atualiza e remove plugins DSH em uma aba da página Plugins. A instalação chama o serviço pluginManager do perfil em processo, permitindo adicionar plugins com o app desktop aberto, sem depender apenas do CLI. Atualizações usam a API pública update API v1 do dsh-market (obrigatória para essa seção) e mostram fase, porcentagem e bytes baixados em uma barra. Só faz rollback quando o market informa um restorepoint. Se a remoção de um bundle em uso for recusada, desativa-o, remove-o do manifest do perfil e agenda os arquivos restantes para limpeza após a próxima inicialização.

Cockpit de desenvolvimento em ciclo fechado integrado ao DeepSeek Harness: em um painel, o agente compila, grava, verifica por depuração e monitora placas STM32; o usuário acompanha tudo ao vivo: etapas de progresso, logs contínuos, uso de RAM/FLASH e formas de onda serial decodificadas. Requer Python 3.10+ do host no PATH (o plugin não pode instalá-lo). O toolchain ARM é baixado automaticamente no primeiro uso (~400 MB, somente Windows x64); requer placa STM32 com debug probe.

Pesquisa literatura médica e de ciências da vida por tema, palavra-chave ou DOI via Europe PMC, PubMed, OpenAlex e Crossref. Verifica acesso aberto pelo Unpaywall/OpenAlex/Europe PMC; se houver uma rota OA, baixa o PDF após validar magic bytes e número de páginas. Caso contrário, cria um cartão de decisão com o motivo exato da falha e quatro alternativas manuais: AbleSci assistance, Sci-Hub, SciDown e X-MOL. A ferramenta Python incluída usa apenas a biblioteca padrão.

Image Designer conecta o DSH a designers da Shenhuo AI, com OAuth pessoal, uploads de materiais, propostas, confirmação da cotação de créditos e downloads locais.

Ferramentas de agente para um downloader Gopeed local: criar, inspecionar, pausar, retomar e remover downloads; diagnosticar ou iniciar o processo Gopeed; receber eventos de conclusão por uma tarefa em segundo plano do DSH que desperta um agente ocioso. Diagnóstico e inicialização do processo são apenas para Windows; as ferramentas de download funcionam onde o Gopeed roda.

Adiciona ao DSH Web GUI uploads em partes com progresso e retomada, além de downloads nativos de arquivos do workspace e do ZIP do log da sessão, inclusive em sessões acessadas por link remoto pareado.

Navegador e exportador de assets Unreal Engine 4/5 (UE4, UE5) para Windows x64: explora, pré-visualiza e exporta assets cooked de arquivos PAK locais via CUE4Parse. Inclui prévia de malhas 3D, reprodução de animações e índice pesquisável. Arquivos comprimidos com Oodle precisam da DLL oo2core da Epic, não incluída, mas disponível para download em Settings.

Plugin de gatilho por eventos (a campainha ao lado do despertador do dsh-proactive): fs watchers recursivos com janelas debounce e filtros glob por regra; endpoint webhook somente loopback com token por regra; regras que despertam uma sessão alvo pelo serviço agents do host (follow-up direto em sessões ativas, retomada primeiro nas inativas) e injetam prompt com variáveis paths/count/payload. Recarrega regras sem reiniciar DSH e registra cada disparo em histórico JSONL auditável, consultável por trigger_list e trigger_history. Rota health loopback informa instalação e estado em runtime. Usos: organizar Downloads, diagnosticar alertas CI/monitoramento de scripts locais e observar erros em logs. JavaScript puro, sem build, MIT.

Pré-verificação de permissões graph-first para DSH. O modelo declara um grafo de computação cujos nós descrevem a operação (o que baixar e para onde, o que gravar ou excluir). O usuário aprova cada nó no painel lateral; uma aprovação cobre uma execução bem-sucedida. Cada nó mantém auditoria do que realmente foi executado. Operações sem cobertura ainda pedem aprovação pelo cartão nativo; um nó bloqueado por permissão sem resposta fica suspenso, enquanto tarefas sem dependência podem continuar.

Adiciona uma ferramenta que captura a área de trabalho do Windows em PNG e entrega o caminho ao modelo, além de um botão de câmera no composer para pré-visualizar, baixar ou copiar a captura.

Navegador integrado ao DSH: Chromium dedicado com perfil persistente para manter logins Feishu ou SSO após reinícios, controlado por CDP. Expõe sete ferramentas de agente (status, open, eval, read, feishu, screenshot, tabs) e espelha a página ao vivo em um painel lateral direito que as chamadas abrem automaticamente. Também converte documentos Feishu em Markdown percorrendo a rolagem virtual. Requer Chrome, Edge, Chromium ou build Chromium do Playwright instalado localmente (não baixa nenhum); defina o binário em DSH_IAB_CHROME.

Baixa vídeos e cria GIFs a partir de links ou clipes locais no Windows x64, com limite de tamanho opcional.

Ambiente de trabalho com dispositivos HarmonyOS para DSH NEXT Desktop no Windows x64 (testado com DSH NEXT 2.0.21-next e Better Sidebar 0.28.2). O espelho flutuante reutiliza o Electron do DSH ou baixa cerca de 140 MB no primeiro uso.