DeepSeek Harness Plugins · Model Inference
Plugins that extend DeepSeek Harness (DSH) with vision, workflows, memory, web tools and more.
Provides model-driven context compression tools for DeepSeek Harness, letting the model decide when and what to compress into high-fidelity summaries to reclaim context.
Deployment-configured prompt prefix: ordered system-prompt sections render ahead of the persona, and reference user/assistant exchanges are injected into every ordinary request—zero log writes, byte-identical prefix reuse.
Turn long articles, books, PDFs, and document sets into traceable claims, evidence, confidence levels, knowledge maps, and review questions with five reading modes.
Adds a dedicated Claude provider to DeepSeek Harness with reasoning settings, model discovery, and known-model defaults.
Adds image reading, document conversion, and local image editing to text-only models, with native thumbnails, OCR, pixel analysis, and optional external vision APIs.
Adds a three-column workbench to DeepSeek Harness with an agent control plane, smart terminal, files, Git, usage, and Ultra Slash commands.
A DeepSeek Harness plugin that adds file uploads, native image input for vision models, and multi-format document reading.
Antigravity access for DeepSeek Harness with OAuth login, multi-account pooling, 429 rotation, device fingerprinting, and CLI/web management.
A local guard for DeepSeek Harness academic writing that checks style, evidence integrity, journal fit, and delivery cleanliness.
Chat with a Harness Agent from Feishu or Lark via WebSocket long connection, reusing models, tools, and sessions.
Lets DeepSeek Harness agents drive ComfyUI to generate images and videos while managing workflows, assets, and job queues.
Use your ChatGPT/Codex subscription directly in DeepSeek Harness without an OpenAI API key, with models, quota, search, and image generation built in.