Plugins

One-shot DSH skill (SKILL.md) generator: capture intent, draft, validate, package and distribute skills from inside a DeepSeek Harness session (skill_new / skill_validate / skill_package / skill_improve_description).

Prompt optimizer for DeepSeek Harness: a composer tool-row button that rewrites the draft into a structured prompt via the session's selected model, with one-click restore

Full-duplex voice mode for DeepSeek Harness: zipformer2 streaming ASR → editable draft, Edge TTS sentence-by-sentence read-aloud with live captions, true barge-in — on-device ASR, no API key. · DSH 语音双工对话：流式识别入草稿、按句朗读+实时字幕、开口即打断，识别本地推理、无需 API Key

Adds a 'file' row to the DSH web GUI composer plus menu: pick images (native composer thumbnails) or arbitrary files (video/audio/Office/PDF, up to 50MB), stored in the DSH home and inserted into the draft as native file reference chips the agent can read

Persistent draft sessions for DeepSeek Harness — prepare multiple independent tasks and send them later.

Long-form drafts with the structure already decided. Blog posts, press releases, product and job descriptions, FAQs, email templates, READMEs, changelogs, meeting minutes, weekly reports, cover letters, academic polish, name ideas.

Reusable prompt presets for DeepSeek Harness: manage a preset library in Settings and inject any preset at the start of the composer draft with one click. / DeepSeek Harness 提示词预设库：设置页管理、对话框一键注入。

Rewrite the current composer draft through an auxiliary LLM call without sending a message.

Check the dsh plugins you already installed against a registry of verified packages: which ship a broken package, which are pinned to a dsh version that no longer exists, and which run code on your machine at install time. Drafts the issue that fixes each

Prompt optimizer in the composer: rewrites the draft into an eleven-section analytical prompt using the conversation, workspace files and the session's selected model, with one-click undo.

Voice input for the web UI: a mic button in the composer that transcribes speech into the draft via the Web Speech API, with an optional auto-send toggle.

Turns markdown into a WeChat Official Account (微信公众号) draft: inlines every CSS rule onto its element because the editor discards `<style>` blocks and class names, uploads each local image through WeChat's own endpoint so it is not filtered out, and creates the draft; it never broadcasts.

Upload files and folders straight into the conversation — images attach as real message blocks, text content is written into the composer draft, folders land by directory structure.