Ollama now runs DeepSeek Harness locally.
One command and you have integrated web search, without depending on an external server. And the trajectory view shows what the AI does behind the scenes. Real transparency in agents.
Deepseek just open-sourced the layer that actually makes AI agents useful.
And the biggest feature isn’t the model. It’s the architecture most people will scroll straight past. What DeepSeek Harness changes: → Everything is a plugin: model, tools, memory, skills, sandbox,
I made this entire video from my phone.
The AI did the rest on my PC. A DeepSeek agent in PawFlow drove ComfyUI, generated footage with MiniMax H3, checked it visually, added voice-over + subtitles, and assembled the final cut. Self-hosted. Real tools.
Using DeepSeek V4 Flash via OpenCode.
For the prompt, ChatGPT Luna generate a detailed, HTML-based prompt designed to create a snowboarding game.
You have reached the end of the archive
All of deepseek