@qizhen2021/dsh-plugin-vision: A Vision Analysis Tool for DeepSeek Harness
DeepSeek Harness (DSH) enables text-only models to gain visual processing capabilities through the plugin `@qizhen2021/dsh-plugin-vision`. The plugin registers a `see` tool that provides three analysis channels: offline OCR (supporting macOS language detection), ASCII layout maps, and semantic descriptions (based on the `mimo-v2.5` model). Installation requires modifying `package.json` via Profile Bundles and configuring parameters such as the key path and timeout in `cordis.patch.yml`. When invoking the `see` tool, parameters can be specified to select the processing channel, and a structured JSON report is returned. Note that OCR is limited to macOS, large images require preprocessing, and the plugin's first
Read Moredsh-vision-suite: A Windows Plugin to Supplement DeepSeek Harness with Vision Capabilities
# Foreword When building agents with DeepSeek Harness (DSH), relying on DeepSeek's text model for reasoning is fine, but many tasks require "seeing": reading error messages from screenshots, comparing layout differences between versions, or extracting primary colors from design drafts. The common practice is to manually send the image to another multimodal service and then transfer the results back to the conversation, breaking the workflow in the middle. dsh-vision-suite addresses exactly this issue. While retaining the official `deepseek-official` text routing, it integrates image understanding, OCR, screenshot cropping, pixel comparison, primary color extraction, and secure webpage screenshots into Harness. Text reasoning remains handled by DeepSeek; the plugin will only forward...
Read Moredsh-img: Image understanding and local visual tools for DeepSeek Harness
dsh-img is a community visual plugin for DeepSeek Harness designed to provide image understanding capabilities to pure text models. By integrating with a vision backend or local pixel tools, it enables `analyze_image` for image Q&A, OCR, layout understanding, and `vision_ground` for element localization. The plugin supports Zhipu GLM-4V-Flash, Tongyi Qwen-VL, Ollama local models, and OpenAI-compatible endpoints, and includes key-free local tools such as `vision_crop` and `vision_ocr`. Installation requires Node.js >= 20 and the DeepSeek Harness core, and it is installed into a specific profile via `pnpm add`. When using an API backend, the API key must be injected.
Read MoreDSH Computer Use for Windows: Windows Desktop Operation Tools for DeepSeek Harness
`Altairpaca/dsh-computer-use-windows` is a Windows desktop automation plugin for DeepSeek Harness (DSH), currently in the experimental Alpha phase. The project aims to address challenges in Windows automation, such as irrelevant windows appearing in screenshots, unstable coordinate control, and OCR text drift. Core features include window-bound screenshots, robust OCR, post-click verification, mouse and keyboard input control, window management, batch execution entry points, optional vision model integration, and local health checks. The system supports a pure OCR mode to keep screenshots local and allows configuration of OpenAI-compatible vision providers. Requirements for installation and execution are Windows 11, PowerShell 7.4+, and Node.js.
Read Moredsh-eye-vision: Image Understanding, OCR, and UI Analysis for Text-Only DSH Models
dsh-eye-vision is a DSH plugin for DeepSeek Harness, designed to address the pain point where the main model only supports text input but needs to process image tasks. This plugin achieves image understanding, OCR, and UI analysis by invoking arbitrary OpenAI-compatible multimodal APIs, returning the results to the main model as text without requiring the main model to directly support image input. Its core features include providing the `image_understand` tool, support for large image cropping, directory whitelist configuration, hot-reload functionality, and stripping proxy environment variables. Users can quickly install it via the `dsh plugin add` command, which supports local paths, URLs, or data URIs as image sources, and allows for automatic recognition by specifying `task_type`,
Read MoreDSH DeepSeek Vision: Adding Image Understanding, OCR, and Visual Evidence to the Text DSH Model
DSH DeepSeek Vision is an MIT-licensed plugin maintained by Argonaut790, designed to add vision capabilities to the text-only DeepSeek Harness model. It introduces an independent vision routing system, providing image understanding, full-screen OCR, and persistent visual evidence features, enabling text-based models to invoke image analysis capabilities and retain results within conversations for traceable evidence. Core features include support for multiple image selection strategies, a session-scoped vision analyzer, structured summary output, and a global Vision provider selector. Installation requires specific Node.js and Harness version requirements, is not published to npm, and must be built locally and added to a Web profile, with configuration needing to avoid conflicts with built-in vision tools. This plugin is suitable for scenarios where images need to be stably integrated into
Read MoreDS-Vision-Plugin: A Web Image-to-Text Plugin for DeepSeek Harness
ds-vision-plugin is an open-source plugin designed for DeepSeek Harness, aimed at addressing the issue that text-based models cannot directly process image inputs. The plugin automatically converts pasted or dragged images into text within the web interface for DeepSeek models to continue reasoning, eliminating the need for users to manually switch tools or models. Its core features include support for a four-model racing mechanism (such as the agnes and glm series) to achieve fast and reliable first valid responses; routing support for Baidu OCR or local Tesseract, with fallback to a vision-language model when unavailable; and the ability to integrate custom OpenAI-compatible models or local runtimes (Ollama/LM Studio). Additionally, the plugin offers comprehensive diagnostic tools and YAML hot-reload functionality.
Read More