DeepSeek Harness Plugins · Client
Plugins that extend DeepSeek Harness (DSH) with vision, workflows, memory, web tools and more.
dsh-plugin-diraud enhances DeepSeek Harness plugin management by grouping the plugin list by source, allowing users to distinguish official plugins from self-installed ones.
Displays a live USD cost badge for the conversation in the Web UI, including per-turn cost and session totals.
This plugin provides audio and video preview for the DeepSeek Harness web GUI, integrated as a FileViewer in the sidebar with streaming playback and HTTP Range support.
A DeepSeek Harness plugin that monitors files created or downloaded by the agent during a session and displays them in a right-side popup panel with preview, open, and copy path options.
A DeepSeek Harness Web plugin that adds a real interactive terminal with a VS Code-style collapsible panel, running shell in PTY and using xterm.js for full terminal rendering.
A workspace file-explorer plugin for DeepSeek Harness Web GUI that adds a VS Code-style panel with file tree and editor for AI Agent collaboration.
Adds Hong Kong Traditional Chinese support to DeepSeek Harness by converting Simplified Chinese to Traditional with localized phrasing, keeping product terms in English.
A real-time billing monitoring plugin for DeepSeek Harness Web GUI that tracks model usage and costs, with support for balance anchoring, official price synchronization, and ledger persistence.
A client theme plugin that transforms the DeepSeek Harness interface into a Cyberpunk 2077 Night City style, featuring neon colors, atmospheric effects, and customizable settings.
A DeepSeek Harness plugin that displays an iOS-style black bar below the input box, expanding to a panel showing session cost, account balance, cache hit rate, token usage, and other statistics.
DSH NP-PPT is a full-featured Presentation Expert plugin for DSH, offering PPTD DSL, WYSIWYG editing, and high-fidelity compilation for one-click PPTX export.
Adds a microphone button to the DeepSeek Harness Web UI for real-time voice-to-text input, using the browser's Web Speech API with multi-language support.