Plugin

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.

Reads assistant replies aloud via Fish Audio API only (bring your own key): per-message read-aloud, auto-read toggle, and a settings page for model, voice reference_id, encrypted API key, and proxy.

Full-duplex voice mode for the Web UI: a composer mic (RMS endpoint detection) transcribes speech with whisper running locally in the browser, assistant replies stream back as spoken audio sentence-by-sentence, and speaking interrupts playback and the running turn (true barge-in). No API key.

Sound reminders with one-click jump for concurrent sessions: the current session gets crisp dang/dang-dang tones, other sessions a soft ding/ding-ding plus a top-right card that jumps straight to the replying conversation.

Voice AI girlfriend for the Web UI: FunASR mic input, Qwen3-TTS spoken replies, companion animation window, and two-way QQ chat (text/voice/image push) via NapCat.

Give text-only models vision: a describe_image tool that delegates images to a vision model from your dsh model list over the harness's own LLM runtime.

Vision for any DSH route: paste images in the Web composer with intent-aware auto-analysis, delegate workspace image reads to a Kimi/MiniMax vision subagent, and materialize pasted originals for editing.

Hands repetitive text and vision labor (OCR, image analysis, comparison) to a locally running Unsloth Desktop (Unsloth Studio) server through unsloth_run and unsloth_vision tools; pure HTTP client, never spawns or owns processes.

Give your text-only model eyes - chat image attachments are auto-described via a vision model (default prompt), with iterative re-parsing through model-generated prompts when details are missing; system/custom model modes + GUI config panel, key-safe secret handling, and a small host patch for DSH 0.1.0-rc.6 (see repo README).

Vision provider route that transcribes attached images to text through a configurable model (15+ OpenAI-compatible and Anthropic vendors) while DeepSeek keeps answering.

Two-tier image reading for text-only models: fast local OCR (RapidOCR, offline) first, vision-model fallback (modlens).

Transparent image guard for text-only routes: paste images without the 400 session deadlock, plus a vision_analyze tool for OCR/PDF/docx/pptx/video.

Image-to-text input for the Web UI: paste or drag an image and it is transcribed into structured text and sent, giving text-only LLMs image-input takeover (OpenAI-compatible vision API).

Give DSH text-only models vision: an image/OCR/document recognition skill (race pool → custom channels → local) plus an idempotent host patch so image messages reach the model.

Headless-browser verification tools so the agent opens the page it just built, reads the rendered DOM and computed styles, checks the console, and screenshots the result; ships a frontend-verify skill.

Browser operation by accessibility refs: navigate, snapshot the page as a ref tree, act on elements, and wait for conditions; the origin policy follows the session approval stance and is enforced by request interception.

Chrome side panel that embeds the dsh web UI and lets the agent read the current page, capture HTTP traffic, and drive the browser.

Keep the built-in web_search tool without an official DeepSeek API key: disable web-search-deepseek and use Tavily keyless by default.

SearXNG-backed web_search provider for the web seam: free, self-hosted, key-less metasearch through your own instance's JSON API, with a bundled loopback-only docker-compose example.

Real-browser automation for dsh via an MCP server (Playwright core): navigate, click, fill, type, screenshot, and act on accessibility snapshots with numbered references - all 22 web_* tools re-exposed to the harness over stdio.

Drive Jiey Browser from DeepSeek Harness over MCP: navigate, snapshot, act, read, screenshot, and manage tabs in a real Chromium profile.

Independent browser acceptance testing for agent deliverables: JSON spec in, real-Chromium verdict out.

Render Markdown + JSON into self-contained HTML documents, cards, dashboards, and galleries (CLI + artifact_render tool).

Human-in-the-loop tools (`human_task` / `human_task_ready_check`): pause the agent until the user performs a real-world action or returns an observation (GUI operation, game testing, visual verification, hardware checks), gated by session consent and an AFK presence check, returning a structured JSON result via a web task dialog.