Плагины

Подключает DeepSeek Harness к Postiz через MCP: список подключённых соцсетевых каналов, получение правил публикации для каждой платформы, планирование, создание черновиков или публикация постов в X, LinkedIn, Instagram, Facebook, Threads, TikTok, YouTube, Reddit, Bluesky, Mastodon, Discord, Slack, Telegram и других; добавляет навык рабочего процесса postiz.

Skills и необязательные MCP-строки для одиннадцати маркетплейсов: сравнение цен по Wildberries, Детскому Миру и Яндекс Маркету, плюс поиск по источникам, карточки товаров и отзывы, включая Ozon, Avito и AliExpress. 14 skills загружаются при установке; обе MCP-строки остаются выключенными, пока RU_MARKETPLACE_MCP_DIR не укажет на локальный клон, которому нужны Python 3.12+ и uv.

Второй мозг для вашего агента DeepSeek Harness — проверка мощной моделью перед рискованными вызовами инструментов, плюс 67 моделей из одного кошелька. Без аккаунтов, без API-ключей.

Плагин DeepSeek Harness, который блокирует рискованные установки npm с помощью метаданных и глубокого сканирования цепочки поставок.

Скин Sky: Children of the Light для веб-интерфейса dsh — светлые дети на сияющей звёздной карте, золото свечного света и звёздная ночь.

Kaze Tachinu (The Wind Rises) skin for the DSH Web GUI: a full-screen film wallpaper behind glassmorphism surfaces, sky-blue interaction tokens, redrawn vector logos for the sidebar and home hero, a restyled composer card with themed placeholder text, and a bottom fade on the message list.

Falsification-driven correction loop for agent sessions: risky work commits an evidence-checkable claim first, deterministic gates verify it, falsified claims force mutually exclusive replacement hypotheses with discriminating experiments, and every event lands in an append-only evidence ledger.

Provenance-aware execution security for DeepSeek Harness. Track sensitive data across tool calls and stop risky side effects before execution.

Update copilot for DeepSeek Harness: one scan covers the DeepSeek Harness core and every profile plugin; explains what changed and how risky each update is, then updates only what you confirm.

Computer Use for macOS desktop applications: reads any app's accessibility tree as text with element indices, then clicks, types, pastes, scrolls, drags, selects text and sets values through it, plus screenshots read back as images. Driven by a persistent JavaScript REPL — the agent gets two tools (`js`, `js_reset`) and calls an `sky` API inside them, so state survives between calls and the tool schema stays fixed as capabilities grow. Accessibility trees come back as diffs after the first read; actions address elements by index and indices from a stale read are refused rather than misapplied; typing waits for keyboard focus and reads the field back; paste restores the clipboard. Requires macOS 14.4+, Accessibility permission (Screen Recording only for screenshots) and Node 20+; ships a prebuilt native module, so nothing compiles at install time.

Semantic risk grading and progressive authorization: classifies tool calls into safe/risky/redline, asks before irreversible actions, auto-allows only approved-and-succeeded signatures.

Local and remote SSH terminals for DeepSeek Harness Web with three permission levels, human approval for risky AI commands, and encrypted credential storage.

Codex-style auto-review permission mode: adds an auto-review preset that auto-approves safe sandbox escalations, asks on risky or ambiguous ones, and rejects critical unconfirmed operations.