Plugins

Um enxame matricial de subagentes que roteia tarefas heterogêneas para o modelo mais adequado a partir de um gateway tipo OpenRouter e cfgpu.com/llm/square, despacha cada uma via subagentes em processo ou chamadas diretas de LLM, e rastreia o consumo de tokens por modelo com uma classificação orientada por feedback.

Linhagem de sessão imutável no estilo Git para o DeepSeek Harness

Geração estável e editável de PPTX para agentes de IA — entrada em IR semântica, saída em DrawingML nativo

Portões de verificação com computação real: simplificação matemática, provas lógicas, validação de regex, FOL via E-prover, verificações de máquina de estados e reparo de código, apoiados em binários compilados em Go com executáveis pré-compilados para Windows

Cria pacotes de base de conhecimento auditáveis (SQLite FTS5) a partir de md/txt/docx/pdf, com recuperação determinística e leitura do texto original

Drops a bowl of white rice from the top of the DSH window for every N tokens you spend; the bowls fall and bounce under gravity, can be dragged and thrown, and a vortex button (with a sound effect) inhales them all — or just the one you drop on it. Threshold, bowl size, on-screen cap, auto-righting and artwork are adjustable in an in-app panel.

WeChat (personal WeChat, iLink Bot API) channel: chat with your DSH agent from WeChat, with images, files and voice-note transcription. One WeChat conversation can hold several DSH Sessions (/new, /list, /use) and every option is editable from the settings page. Also published to npm as dsh-plugin-weixin.

Round-table meetings for DeepSeek Harness 0.2.0-rc.2 on Windows desktop, reusing existing Agent sessions with discussion, editable task cards, sequential and parallel steps, shared meeting files, and secretary minutes.

One-button prompt optimizer: rewrites the composer draft via a configurable strategy, with an editable template, few-shot examples, pinned models and reasoning effort.

Generate images from chat by calling a fleet of local or LAN ComfyUI workers, with workflow templates, worker health checks and queue control.

A research library panel and skill set for DSH: project folders, figures with registered provenance, drafts, references, data tables, journal profiles and submissions, plus writing, citation and Origin-plotting discipline for research agents.

Markdown Studio：在 DSH 原生右侧文件预览与对话消息中渲染 Mermaid 图表、KaTeX 公式、GFM 表格、任务列表、脚注及代码高亮。渲染引擎随包内置，完全离线可用。 / Native DSH right-sidebar Markdown preview with Mermaid diagrams, KaTeX math, GFM tables and syntax highlighting — fully offline.

Adds a human-editable notes workspace with folders, full-text search, rich text, checklists, resizable tables, attachments, exports, and opt-in agent collaboration.

DeepSeek Harness plugin: label several directories inside one workspace and reference them by absolute path. Type @ in the composer to pick a label, manage them from the workspace row or the workspace_dirs tool, and the label table is injected into every

Counts completed compactions per session in a floating window and warns only after repeated compactions plus extra token spend or high context occupancy, with all three thresholds adjustable in Settings.

Multi-workspace manager: bind a docs anchor plus ordered code-project repositories into one session, injecting their absolute paths, a recursive file index, a server/client endpoint index and optional coding standards, and syncing writable directories into the sandbox.

A dsh-better-sidebar reader for CSV/TSV/PSV files and xlsx/xlsm workbooks: a sortable, filterable table with sheet tabs. Ceilings on rows, columns and cells, plus two zip-container gates, keep an input from freezing the tab.

A dsh-better-sidebar reading view for .docx files: headings from the style map, paragraphs, lists, real table cells, inline pictures and header notes. Ceilings on archive size, per-part inflated bytes and rendered blocks keep a document from freezing the tab.

Read-only, scriptable session playback for the DeepSeek Harness WebUI

Texas Hold'em table for DeepSeek Harness: model-callable poker tools, six heuristic opponents, an optional coach, and a browser table whose buttons never post a chat message.

Task-to-skill pairing meta-skill with a laziness ladder for minimal loading, a user-editable routing table, and a quarantine, SkillSpector scan, human-approval install flow.

Turn-rail settings for the DSH Web conversation locator: tick thickness, rail side, a curved length gradient anchored on the hovered tick, and a plain-text hover preview with adjustable line count, font size, and width.

Email notifications for DSH with an away-mode switch: mails you when a turn ends, a tool waits for approval, or the assistant asks a question. Title and body are editable templates.

TypeSafe Jev (modelo de decisão System One) para o DeepSeek Harness, de ponta a ponta: jev_decision (choice/score/noul, paralelo, respostas tipadas + confiança), um cartão em Settings > Plugins para a API key e chaves de liga/desliga, visualizações inline de ferramenta mostrando cada decisão na conversa (resposta, confiança, latência, custo), um auto-guard (checagens determinísticas + de risco/loop do Jev), um dashboard local /jev e um benchmark de verificação online (96,3% em 27 perguntas rotuladas, mediana ~283 ms no jev-latest, 2026-09-21). Honesto por design: só API real, sem fallback simulado.