DeepSeek Harness Plugins
Plugins that extend DeepSeek Harness (DSH) with vision, workflows, memory, web tools and more.
Render editable and exportable AntV Infographic SVGs directly inside DeepSeek Harness (DSH) replies.
dsh-seelog is a DeepSeek Harness web plugin that converts Session Logs into a browsable session execution map with a main agent thread and sub-agent forks for easy navigation and traceability.
A Git-backed long-term memory and evolution plugin for DeepSeek Harness that stores session notes, branch records, and skill drafts in a Git repository for persistence and versioning.
dsh-reconnect is a plugin for DeepSeek Harness that safely retries failed model requests with exponential backoff, recovers from relay/proxy failures, and includes a visual configuration panel.
A stateful, Agent-controllable Jupyter workspace for DeepSeek Harness, providing a persistent kernel and tools for agents to read, edit, execute, and inspect notebooks.
dsh-plugin-manager is a third-party plugin manager for DeepSeek Harness, allowing users to search and toggle plugins on or off in the Web UI settings with a single click, taking effect after restart.
A UI skin plugin for DeepSeek Harness that adds wallpaper or video backgrounds with a frosted glass effect, supporting drag, crop, and history records.
The dsh-guandan-panel plugin embeds the Guandan card game into the DSH interface as a bottom panel, offering single-player gameplay with AI opponents, keyboard shortcuts, and a table overlay.
A plugin for DeepSeek Harness Web interface that monitors LLM quota and usage in real-time, displaying a quota badge and providing detailed usage statistics.
An LLM adapter plugin for DeepSeek Harness that accesses WorkBuddy/CodeBuddy models via a local proxy, requiring no API key, and includes a web status widget.
A dsh plugin implementing Best-of-N self-verification to enhance LLM inference quality like frontier models at a fraction of the cost.
A free multi-engine web search and fetch plugin for DeepSeek Harness, featuring automatic fallback and dedicated search endpoints with zero LLM token costs.