Plugins

Adds a Screen-sense submenu (turn on, turn off, and an interval of 15, 30, 60 or 120 seconds) to the dsh-pet desktop right-click menu, forwarding each choice to dafeiyu-screensense. dsh-pet exposes no menu extension point, so this plugin appends a marked patch to its renderer.js after backing the original up to renderer.js.dafeiyu.bak; the patch is idempotent and is re-applied when dsh-pet is reinstalled.

Locks the model's reasoning, replies and tool-call descriptions to a chosen target language (Simplified Chinese by default) with system-prompt sections at the head, middle and tail of the prompt, offering three strength levels, an auto mode that follows the DSH interface language, a settings panel that shows the language currently in effect, and a sidebar quick switch.

GitHub project search and analysis: describe your idea in one sentence and the agent searches GitHub for related projects, ranks them by relevance weighted with stars, and returns a structured card — last updated, relevance, stars, weighted score, language and a one-line description for each project. Pick a few repositories and it reads their README, dependency manifest, directory tree and key files to lay out their architecture — so you stand on the shoulders of giants before you start building.

Transforme un lien Xiaohongshu, Douyin, TikTok, YouTube ou X en texte : transcription avec codes temporels, texte à l'écran, description et OCR de chaque image, légende et métadonnées. Monte le serveur MCP hébergé via streamable HTTP.

Un compagnon en pixels qui vit dans la zone de chat : éclot d'un œuf en l'un de trois sprites ou en un sprite personnalisé dessiné par le LLM à partir d'une description ; saute tout seul par-dessus les bulles de messages, se contrôle au WASD, discute en personnage via les modèles que vous avez déjà configurés, et embarque un système d'élevage (humeur/énergie/santé, niveaux, boutique) alimenté par l'usage réel de jetons.

Pièces jointes d'image natives pour DeepSeek en mode texte seul dans la Web GUI : les images collées ou déposées apparaissent sous forme de vignettes dans la session, et avant l'envoi l'hôte les lit avec l'API de vision gratuite Zhipu GLM-4V-Flash (chaîne de repli glm-4v-flash) et substitue la description, de sorte que DeepSeek répond sur l'image tandis que l'original est conservé dans l'historique.

Récupère le mode vision intégré de DeepSeek web pour les modèles texte seul : l'outil deepseek_vision pilote l'automatisation de navigateur locale deepseek-vision-cli (aide à la connexion manuelle, deep-think activé, fermeture automatique du navigateur) et renvoie les descriptions d'images sous forme de texte.

给 DeepSeek Harness 工具加上模型填写的步骤说明

Search field on the DeepSeek Harness Plugins page that filters the installed and official plugin cards as you type, matching each card title, description and package name; space-separated keywords must all match.

Find and compose DSH plugins from a task description, then open an independent chat environment or install the plan into the current profile.

Ships a skill about writing DSH skills - SKILL.md frontmatter and root precedence, how to phrase a description so it actually triggers, length and splitting rules, and an A/B check that the finished skill changes agent behaviour. The package also carries the checker the skill uses to validate a skill directory.

Searches a database of past motor-repair cases by fault description, motor rating and brand, and lets new diagnoses be saved back for future retrieval.

Three independent language channels for model output — tool-call descriptions (every tool-call card title, including run_code in PTC-style presets), thinking, and replies — each set to follow the web GUI language, force a specific language, or turn off (default: only descriptions, following the GUI). One global runtime-context prompt directive on the host; the client half reports the GUI language; a Settings plugin card switches the modes.

Transforme un besoin énoncé en choix de plugin : une skill d'agent qui interroge un catalogue du sujet `dsh-plugin` entier (2600+ repos, recrawlé toutes les 6 heures, descriptions chinoises depuis Workers AI) via une API de recherche classée, puis rend la commande `dsh plugin add` correspondante.

Outil `describe_image` : un pont de vision qui envoie les images à mimo-v2.5 via l'API opencode Zen (identifiant `OPENCODE_GO_API_KEY`, route gratuite d'abord avec repli payant) et renvoie des descriptions texte aux modèles texte seul, avec passe-plat natif et transcodage ImageMagick des formats SVG/TIFF/HEIC.

Pont de vision à découverte automatique pour les agents DeepSeek Harness texte seul : trouve automatiquement un modèle capable de traiter les images parmi vos fournisseurs configurés et renvoie les descriptions en texte brut via un outil de vision.

Adaptateur DeepSeek augmenté par la vision : un modèle capable de vision décrit l'image en entrée, puis un modèle DeepSeek texte seul raisonne sur cette description.

Montez une seule ligne dans la composition et chaque carte de plugin de la page de réglages web reçoit une description bilingue (zh/en) ; publie aussi le service pluginDescriptions pour que d'autres plugins puissent enregistrer leurs propres descriptions.

Bundles the 32 gis/ skills of opengis-skills (GDAL, QGIS, GeoServer, PostGIS, PROJ, PDAL, GRASS GIS, JTS, GeoPandas, Shapely, CesiumJS, OpenLayers, Leaflet, deck.gl and more) into the DSH skill catalog; only names and descriptions appear up front, and instruction bodies plus reference documents load on demand.

Shows what each installed skill costs in the model context and hides a skill from the catalog without deleting it. Lists every skill found on disk with the exact catalog line the model receives (description normalized and truncated at the catalog cap), its body size and token estimate; previews the catalog message byte-for-byte as the agent will receive it, compares that with the copy already published in the current session, and reports skills on disk that never reach the catalog (invalid frontmatter, or shadowed by a higher-priority root). Hiding writes disable-model-invocation into frontmatter and keeps the file, so the skill still works via /name. Also imports skills from ClawHub, Skills.sh and GitHub URLs and uninstalls user-owned ones.

Sorts every loaded skill into one or more user-owned workspaces in a right-sidebar tab: each workspace is its own category list plus where every skill sits, categories are created, renamed, reordered and deleted in place, and skills are dragged or picked between them. A first run can let the profile model build the first classification, or place skills into categories the user built first; clicking a row copies its /skill-name and hovering unfolds the description the skill ships.

Structured image recognition: local sharp measurements plus a cloud vision model return description, OCR with coordinates, object locations and fine-grained attributes in one call; a fingerprint cache makes repeated images return in milliseconds without spending model quota.

Renders installed DSH plugin titles and descriptions in Simplified Chinese: at startup it translates the plugins the local JSON cache lacks through the harness LLM service, then projects { en, zh } metadata for resolveText() to render — no DOM changes, no third-party API.

Tool-surface snapshot and drift gate: baselines the full visible tool registry (name/description/parameters, MCP tools included) into a committable JSON file and warns loudly when an upgrade adds, removes or reshapes a tool.