Plugins

Bringt beliebige Dateien in die Unterhaltung — im Arbeitsbereich abgelegt und per Pfad referenziert; keine Ablehnung nach Dateityp

Globaler Skill-Hub für dsh: bündelt Claude-Code-, Projekt- und .skill-package-Skills in ~/.dsh/skills, mit Import per Referenz/Kopie und Nutzung über das Slash-Menü

Registers bundled full-stack, frontend-design and browser-automation skill documents as four read-only tools; a focus argument selects which reference file is returned alongside the skill body.

Local-first resume warehouse for DeepSeek Harness: package your work into reusable bullets, import an existing resume into a template, analyse a job description (including the school-tier preference it implies), assemble a tailored Chinese or English resume and export a PDF named position-name-school. The whole app ships inside the npm package — server/, web/, skills/, scripts/, requirements.txt, no git clone needed — and one tool starts, stops and health-checks it; the first start installs the dependencies into app/.venv. Needs Python 3.10+ and macOS or Linux (bash). Nothing leaves the machine except the LLM calls you confirm.

After user setup, guides native subagents toward model preferences by task type, using native model-selection permissions with no extra routing-model calls.

Read-only semantic code navigation for DeepSeek Harness via local language servers: definitions, references, symbols, diagnostics, hover, and rename preview.

Skill for building a game design brief into a playable, publishable TapTap Maker project: it picks the browser workspace or local MCP route, extracts the five elements of the brief, writes a first-round prompt for a minimal core loop, iterates in the official order, checks preview and release materials, and keeps credit, credential and irreversible-action limits in view. Includes eight reference notes on the product, credits, themes and asset limits, the leaderboard, brief parsing, iteration, publishing and safety, four templates, and three read-only Python checks for brief completeness, project layout and release materials.

Right-click menu and drag support for the Files sidebar on Windows - copy absolute or workspace-relative paths, create files and folders, rename, move entries to the Recycle Bin, and drop a row into the composer to insert an @ reference.

DeepSeek Harness 插件：把对话整理成论文。按说话人还原上下文、提炼带出处的论点、生成论文骨架，并用 Crossref/DataCite/OpenAlex/arXiv 逐条核验引用。

Reference links panel for DeepSeek Harness: collects every URL of a conversation into the right sidebar with pins, turn counts, metadata and favicon caching. DSH 右栏「参考链接」面板。

Click the chibi mascot and it reads the assistant reply aloud through your local IndexTTS 2.5 install. One character = one reference clip; a missing portrait just shows a placeholder. Ships no voices or artwork — the library lives on the user's own machine. Requires a local IndexTTS 2.5 install.

Deterministic review of synthetic AI-agent claims in DeepSeek Harness, reporting bounded verdicts, reasons, and evidence references.

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_*`.

Cross-session reference cards: drag a Q&A card out of any session into another one and it stays there; when you ask a question in the target session, the dragged cards' question, the answer it got at the time and its attachments are injected into the prompt as earlier conversation context, so the model can quote them directly.

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.

Folder explorer for the DeepSeek Harness right sidebar: open any local folder, browse a tree with file-type icons, rename/create/delete/copy files and folders, and add files to the chat as @ references.

Langzeitgedächtnis über den Besitzer: Extrahiert aus jedem Turn Zeitplan, Vorlieben, Abneigungen, emotionale Auslöser, aktuellen Fokus, Zusagen und Tonpräferenzen und fügt jedem Prompt eine kurze Zusammenfassung hinzu. Meldet sich nach Zeitplan oder mit einem fürsorglichen Satz in Antworten, lernt Aufwach-, Arbeits- und Schlafzeiten aus kurzen Nachrichten und komprimiert Erinnerungen automatisch. Erkennt Plugin-Quellen und ausgehende Werkzeuge, gewichtet ungeprüfte Erinnerungen geringer und kann sie mit einem Klick zurücknehmen.

Rechtsklick auf Dateien, Ordner, Weblinks und Workspaces im DeepSeek-Harness-Chat oder in Seitenleisten, um sie zu referenzieren, kopieren, vorschauen, in einer ausgewählten App oder einem Browser zu öffnen, anzuzeigen oder zu speichern.

Gitee-Enterprise-Arbeitsbereich für Work-Items in der Harness-Seitenleiste: Projekte, Aufgaben, Anforderungen und Bugs durchsuchen; Details, Unteraufgaben, Kommentare und Markdown ansehen; Work-Items erstellen/bearbeiten und mit @ referenzieren.

Care-Engine für DeepSeek Harness, damit sich der Assistent um eine Person kümmert. Wetter als erste Quelle mit gefühlter Temperatur und Luftfeuchte für Kleidung-/Regenschirmtipps, zuschaltbarer täglicher Hinweis passend zur Routine sowie Extremwetterwarnungen und Erinnerungen. Hinweise werden beim Aufwachen des Agenten ausgewertet; ein Host wie qq-bridge kann ihn zeitgesteuert wecken. Einstellungen bleiben lokal.

Übermittelt VS-Code-Auswahl oder Dateipfad als `@path`-Referenz samt Zeilenbereich in den Composer der offenen DSH-Unterhaltung, ohne sie zu senden. Benötigt die im Bundle enthaltene VS-Code-Begleiterweiterung `.vsix` (Ein-Klick-Installation im Composer oder manuell).

Reflexionsschicht für DSH-Sitzungen, die schrittweise ein Weltbild mit 3–7 Einträgen zu Werten, Vorlieben und Gewohnheiten umschreibt, Geist, Gesundheit und Guthaben verfolgt, Bilibili-/Douyin-Videos einliest und ein prüfbares Wachstumsarchiv führt. Videoaufnahme optional, benötigt Site-Cookies; Douyin-Parsing erfordert zusätzlich Edge/Chrome und yt-dlp.

Einstellungskarte für Sessions, deren Verlauf auf Anhänge verweist, die im lokalen Speicher fehlen. Sichert das Protokoll und ersetzt hängende Verweise durch Textplatzhalter, damit die Session wieder senden kann. Reparaturskript benötigt Node 22.15 oder neuer; Karte nur unter Windows getestet.

Öffnet in der rechten Sidebar ein temporäres Nebengespräch mit der Host-Gesprächsoberfläche und gleichem Markdown, Tool-Schritten, Timing und Tokenverbrauch. Markierter Antworttext wartet als Zitat-Pill im Composer, geht mit der nächsten Nachricht mit und wird Kontext des Nebenchats. Eigene, nur speicherinterne Session; der Eltern-Snapshot ist nur Referenz, ins Session-Log wird nichts geschrieben.