Plugins

Runs PowerShell scripts through base64 encoding, so quotes, $, backslashes and CJK survive as written. Same sandbox channel as the built-in pwsh tool.

Drag-and-drop file upload for PDF, Word, Excel and images: dropped files are saved to a local directory and referenced by path, no base64 bloat in the chat.

Fully-local image understanding & OCR via macOS Vision Framework: `ocr_image` (text, table layout + coordinates) and `view_image` (scene, faces, QR) — paste multiple images into the web input box or pass path/URL/base64; images never leave your Mac.

base64/url/hex encoding, common hashes, and UUID generation.

Sidebar panel for Agnes image and video generation. Images: plain text-to-image, or image-to-image and multi-image composition from up to eight reference images chosen from local files or from results already in the panel, at 1K-4K across eight aspect ratios, with six one-click style presets. Video: agnes-video-2.5-flash at 720p, 4-12 seconds, in three modes that the panel picks for you - plain text, keyframe (first frame, last frame, or both) and reference (up to five images) - plus an optional seed; the task is submitted to the host process, so switching panels, reloading the page and restarting the app all resume it instead of losing the job. The prompt is written into each generated PNG or JPEG and shown again when you hover the result, where one click copies it. Browse the output directory (images and videos on separate tabs) with automatic refresh, copy a path, reveal a file in Finder, or download it; the last 24 generations persist across restarts. Images and videos can also be produced from the chat itself through the agnes_image, agnes_video and agnes_video_status tools. Both the image and the video model can be switched from the panel, with a connection test, and local reference or keyframe images are sent as base64 data URIs so no public image host is needed. Needs an Agnes AI API key saved in a local file (default ~/.agnes-key.txt).

DeepSeek Harness sandbox provider backed by MXC (Microsoft eXecution Containers): wraps ctx.sandbox.confine() argv as `wxc-exec --config-base64 <policy> -- <command>` on the BaseContainer backend. One runtime dependency (@microsoft/mxc-sdk, for the wxc-ex

Sidebar playground for SenseNova and Agnes image and video generation, per-provider API keys and default models, local result management, and sgen_* tools for conversations. Each request runs the bundled sgen command-line tool, any reference image or audio you give by local path is uploaded to the provider as base64, API keys live in ~/.sgen/config.json, and the sidebar saves results under ~/sgen unless you pick another folder.

Local media toolbox for DeepSeek Harness: converts files between 27 video, audio and image formats, trims clips, extracts audio or video tracks without re-encoding, exports embedded subtitles to .srt, compresses images with a before/after preview, reads and rewrites metadata, and encodes or decodes Base64 — all through the ffmpeg already installed on the machine, so files never leave it. The same engine is registered as a media_toolbox tool agents can call.

Inlines an image block as base64 only in the one request where it first appears, and marks that same block offloaded on every later request so DSH's own image-offload projection renders it as the placeholder text carrying the read-only local path. An occurrence is keyed by session id, message id, content-tree path and attachment id, and a request is only counted after it produces its first chunk, so retries do not count as sent. Requests carrying a purpose are neither rewritten nor counted, a later read_image of the same file is a new occurrence and inlines once more, and only that request's copy is rewritten, so nothing is written back to the session.

DSH 图片插件：ask_user_choice 图片/图文混合选项（Web GUI 渲染图片选择卡，可放大查看）+ show_images 在回复中内嵌图片（图片与文字混排）。图片来源支持本地路径 / http(s) URL / base64 data URI。纯插件实现，不改核心包。

Developer quick-reference toolbox for DSH: 12 themed categories (195 entries) plus zero-dependency tools — regex tester, JSON formatter, timestamp converter, Base64/URL codec, cron generator and line diff, in a searchable settings panel.

DSH 插件：通过硅基流动(SiliconFlow)视觉大模型识别/分析图片，支持本地文件路径、http(s) 图片 URL 与 base64 data URL。含持久化的粘贴识别面板（web 端）。