@dundunhan
A plugin that gives text-only LLM agents video understanding capabilities, providing tools for metadata probing, scene-aware frame sampling with optional ASR transcription, and time-anchored Q&A using OpenAI-compatible vision models.
@THU-MAIC
OpenMAIC for DeepSeek Harness: classrooms, slides, interactive widgets, and Socratic teaching.
@XCNXNXNX
Portable tavern plugin for DeepSeek Harness, featuring RPG-style SillyTavern V2/V3 character card generation and tavern roleplay chat, with world book, character card JSON/PNG import/export, panel themes, and local music.
@CheshireJCat
DeepSeek Harness plugin for complete Blender 3D modeling, reconstruction, rendering, validation, and export workflows.
@xmutfyh
DeepSeek Harness (DSH) academic writing guard for papers: AI-writing style detection, evidence preservation, journal-fit calibration, manuscript proofreading, writing_audit & automatic checks. Local, zero network, zero LLM.
@609476965
DSH lorebook-driven fiction writer plugin: import Tavern/SillyTavern character cards and world books, save local Markdown settings, generate novel prose referencing the world.
@MJorgin
Free image reading & generation for DeepSeek Harness — paste-image reading, vision model failover, image generation.
@omdsh-dev
Office document tools for DeepSeek Harness: generate, read, and edit spreadsheets (.xlsx), PDFs, presentations (.pptx), and Word documents (.docx).
@HuanLinOTO
Provider-agnostic AIGC HTTP bridge plus infinite canvas and ffmpeg post-processing for DeepSeek Harness; exposes 13 tools including canvas linking, reroll, and media editing.
@omdsh-dev
dsh-gomoku is a Gomoku (five-in-a-row) plugin for DeepSeek Harness that lets you play against AI or watch two AI models compete on a 15x15 board. It supports human-vs-AI and dual-AI modes, separate model and thinking settings for black and white, editable system prompts, pause and manual takeover, and persistent game state in the browser.
@poiuyjie
Automatically converts images to text for pure-text LLMs using a configurable vision model, enabling image understanding without switching the main model.
@Flyvhidbwo
Vision proxy plugin for DeepSeek Harness. Automatically interprets images attached in the GUI using the official deepseek-v4-flash-vision-exp model, translating them into text for DeepSeek to answer. Supports any OpenAI-compatible VLM from DashScope, Zhipu, OpenRouter, etc., and auto-detects local Ollama if no API key is set. Includes a one-step confirmation during installation.
