DeepSeek Harness Plugins · Model Inference
Plugins that extend DeepSeek Harness (DSH) with vision, workflows, memory, web tools and more.
A single-author novel-writing pipeline for quantized small models, featuring a subagent concurrency setting and a complete creation preset with per-chapter review and full-book assembly.
OpenCode Go inference plugin for DeepSeek Harness with per-model protocol routing (Completions, Responses, or Messages), model discovery, and subscription usage.
Connect locally logged-in WorkBuddy models to the DSH model selector with read-only credit overview, model catalog management, and multi-credential switching.
Registers native keyless Kilo Gateway and OpenCode Zen free-model adapters inside DSH with dynamic discovery, tool support, and automatic context reconciliation.
A harness that equips any AI coding agent with 25 connected specialist skills and a lifecycle router covering planning, TDD, verification, security review, and failure recovery.
Bridges locally-authenticated Antigravity CLI headless sessions into DeepSeek Harness as a first-class model provider with streaming generation and tool-call support.
An embedded C/C++ firmware development toolbox with 4 agents and 8 skills covering FreeRTOS, interrupts, Keil MDK builds, and more, compatible across AI coding platforms.
The dsh-tool-json plugin provides secure and efficient JSON structured querying with JMESPath-inspired syntax, eliminating process overhead for agent handling of JSON data.
A DSH plugin for a minimal auto chess game, supporting human vs AI or AI vs AI matches, with each AI decision powered by real model inference.
A plugin that provides selectable vision models for DeepSeek Harness as a composite DeepSeek + Vision provider, where vision models analyze images and DeepSeek handles reasoning and final responses.
mindspace-dsh-local-rag is a local RAG plugin for DeepSeek Harness that provides a search_local_memory tool for model-directed retrieval, fusing vector and BM25+ search.
dsh-mediacrawler is an MCP adapter that integrates DeepSeek Harness with MediaCrawler for supervised data collection across platforms like Xiaohongshu, Douyin, and others.