DeepSeek Harness Plugins · Model Inference
Plugins that extend DeepSeek Harness (DSH) with vision, workflows, memory, web tools and more.
Adds Codex-style server compaction to OpenAI Responses routes with manual, token-pressure, and overflow triggers, retaining recent context and producing portable checkpoints.
A ModLens-style vision bridge plugin that lets a text-only LLM leverage a vision model for structured image evidence via a vision tool or a visual twin model, with zero runtime dependencies.
Injects a system-prompt section that shapes the model's chain of thought to think in English and open every reasoning block with a configurable opening phrase.
Generate images and videos in DSH sessions via OpenAI-compatible APIs, supporting text-to-image, image editing, and video generation with inline preview.
Ships genre-specific format constraints for 15 writing tasks so the model focuses on content while the tool guarantees structure — one call yields a polished blog post, press release, README, or meeting minutes.
14 remote deterministic tools for agents: SQL transpiling, JSON/YAML, text diff, JWT decode, regex testing, QR codes, unit conversion, and page screenshots. No local code runs.
Turn a DSH session trajectory into a redacted HTML, Markdown, or JSON recap report for handoff, teaching, or debugging.
Registers a tools tab in the DSH conversation header to list each tool call with name, arguments, result, and status in dispatch order.
A DeepSeek Harness web plugin that expands all Think lines in one toggle, persists across sessions, and respects manual collapse.
Cross-platform model search with real-time keyword filtering and one-click session switching for DSH multi-provider setups.
One-click natural-language dispatch to DeepSeek Harness native sub-agents, with built-in recipes that deliver real files and evidence cards.
Provides the doubao_ask tool via local Doubao Relay for web search, image generation, multimodal image recognition, and paste-to-path image handling.