Plugins

Flux d'eau continu pour l'état de tour « Deep diving... » de dsh web : trois courants dégradés en parallaxe balayent le texte comme de l'eau qui coule, remplaçant le scintillement à bande unique d'origine. Respecte les thèmes clair/sombre et reduced-motion.

Un protocole structuré d'état de travail pour les transferts d'agents DeepSeek Harness

Plugin DeepSeek Harness installable via NPM qui regroupe le MCP de mémoire persistante Engram et injecte le protocole de mémoire de l'agent parent.

Bundle fournisseur DIT.ai pour DeepSeek Harness : installe 29 modèles OpenAI Chat Completions et 10 modèles Anthropic Messages depuis un DIT_API_KEY, avec configuration de réflexion spécifique au protocole.

Plugin DG-LAB Coyote e-stim piloté par agent et GUI : huit outils `coyote_*` plus un panneau web DSH derrière une enveloppe de sécurité unique (limites souples, limiteur de débit asymétrique, cooldown de session, plafonds de lecture, fail-safe de déconnexion) ; couche auto-stim optionnelle mappant des événements d'agent (appels d'outils, erreurs, fin de tour) vers des impulsions bornées ; protocole socket V3 officiel avec appariage QR, formes d'onde programmables. Adultes uniquement.

Protocole de collaboration multi-agents indépendant du projet pour DeepSeek Harness : gardes de boucle qui arrêtent les boucles d'agents emballées, format de passation à six champs, routage des revues par niveau de risque et séquence fixe verify→commit→push→review→close. Une seule skill au lieu des mêmes règles copiées dans les instructions de chaque agent.

Bundle ACP DeepSeek Harness pour la délégation à l'éditeur : un profil de serveur stdio Agent Client Protocol en une commande, qui partage les identifiants et journaux de session dsh.

Bundle DeepSeek Harness pour les outils et skills de gouvernance Amber Protocol.

A glass-material appearance for DSH: translucent composer, dialogs, menus and code blocks — one switch in Settings, off restores the stock look exactly.

A browser that lives inside the DSH host: every session gets its own tab, the AI drives it with lazily published browser_embedded_* tools, and a human watches or takes over that same tab in an on-demand right-Sidebar tab — no GUI, no external browser, log

Connects the Xiaozhi voice assistant to DSH Web over MCP: DSH is the tool provider, weaving 35 DSH Web endpoints into 16 voice-friendly tools for workspaces, sessions, chat, models, settings and files — outbound WebSocket to the Xiaozhi MCP access point by default (no public IP or port forwarding needed), several devices bound at once, plus a DSH settings page with live status.

Universal MCP Apps Host & UI Plugin for DeepSeek Harness (DSH)

Browser automation connector: drives the local Edge or Chrome over the DevTools Protocol with zero runtime dependencies — launch a headless throwaway-profile browser or attach to one started with --remote-debugging-port, then navigate, read page text and interactive elements, click, type, and screenshot through 11 agent tools.

Declares the OpenCode Go and Command Code route wire protocol and endpoint (so the API address never has to be typed), adds the DeepSeek V4.1 models to the OpenCode Go route and puts them first, completes the Command Code model list from its public catalog, and attaches a per-conversation x-opencode-session header to OpenCode requests (fixes 400 MissingSessionID).

Connects DeepSeek Harness to the QianYuan MCP server, a shared record of pitfalls other agents already hit and results they already produced, plus capability lookup. Exposes 13 tools named mcp__qianyuan__*. Reads are open and need no key.

Tiered compaction hub: replaces the single 80% compaction cliff with a configurable ladder (default 30/50/70/90) whose tiers escalate from gentle to maximum consolidation laws, folds a deterministic keyword index into every checkpoint so facts accumulate across cycles, and summarizes through a model chain — primary plus up to 3 fallbacks, each with its own reasoning effort. Adds a Compaction module sub-tab and a Settings card; toggling off always restores stock compaction.

OpenCode Go provider that verifies a model against the gateway rather than treating a recommendation as a measurement: the models.dev protocol suggestion is probed before it is used, reasoning levels follow a measured contract, and requests run through an ordered protocol fallback chain. Context window, max output and image-input numbers are declared by a runtime-refreshed models.dev layer rather than probed, with operator overrides above both. Several OpenCode Go credentials can be stored and switched from the settings page.

Declares the wire protocol for the opencode-go route so models the installed catalog does not describe resolve, adds the DeepSeek V4.1 models to that route when they are missing, and attaches a per-conversation x-opencode-session header to OpenCode requests.

Completes @path references in the Web composer: type @ (optionally ./src/ma) to list matching files and directories, pick one with the arrow keys, and the draft receives an @./relative/path reference (quoted when the path has spaces).

A browser that lives in the DSH host: every session gets its own tab, the AI drives it, and you watch or take over that same tab inside the DSH Web UI — logins, 2FA, QR codes and CAPTCHAs included.

DeepSeek Harness plugin for verified competitive-programming problem-setting workflows.

Community ACP bridge for DeepSeek Harness that supports session/load by replaying a persisted session's transcript

Fantasy glassmorphism skin for the DSH Web UI: a moonlit nether-forest scene with drifting teal clouds, red spider lilies spreading from the lower-left and a sleeping cat in the corner, drawn over a WebGL2 fluid background with its own wallpaper and video backgrounds. Ships two switchable variants (pale / saturated blue), three bundled background tracks plus uploads kept in local IndexedDB, and cat-paw, fish, bubble and starlight decorations. One switch restores the stock interface exactly.

Project-based learning coach: a subject-agnostic teaching engine plus swappable domain packs (Unity/C# included), shipped both as a DSH skill and as an installable bundle. The skill protocol enforces attempt-first teaching and graded hints; a zero-dependency validator rejects ability claims their evidence does not support and blocks stage acceptance while blockers are open.