DeepSeek Harness Plugins · Model Inference
Plugins that extend DeepSeek Harness (DSH) with vision, workflows, memory, web tools and more.
An LLM provider plugin that adds NewAPI to DeepSeek Harness without modifying the core, implementing an OpenAI-compatible gateway with model discovery.
dsh-btw is a bypass Q&A plugin that answers questions in an independent read-only session branch via the /btw command, without polluting the main conversation context, supporting multi-turn dialogue and main conversation history sync.
DSH-Office is a DeepSeek Harness plugin that uses the local zagens-office tool to generate and edit office documents such as PPTX, DOCX, XLSX, and PDF.
dsh-claude-cli integrates the Claude Code CLI as a model inference backend for DeepSeek Harness, running it as a subprocess without an API key and streaming output while the harness remains the agent.
A finance data toolkit for DeepSeek Harness agents that ensures accurate number formatting, return math, ratios, time value, and risk metrics; the model explains, and the plugin handles the arithmetic.
A DSH bundle that integrates ForgeaX game development into DeepSeek Harness, providing MCP tools, skills, and a workspace guard.
Fully-local image understanding via macOS Vision, including OCR, table extraction, and description, with images never leaving your Mac.
Adds visual perception to text-only models in DeepSeek Harness via a free or custom vision API, generating and persisting descriptions without replacing the reasoning model.
dsh-calculator is a DeepSeek Harness plugin that displays API spend and account balance in a floating card on the Web GUI, supporting peak/off-peak pricing and real-time updates.
ds-vision-plugin adds image input support to text-only DeepSeek models by automatically converting pasted images to text via a race among four vision models.
This plugin adds visual capabilities to text-only models in DeepSeek Harness by using free Gemini and GLM vision APIs for image description, OCR, and visual Q&A.
dsh-llm-fallback is a DeepSeek Harness plugin that provides automatic fallback to alternative providers when the primary one fails, ensuring uninterrupted model inference.