Agent: CursorLLM: DeepSeek, Claude 3.5#jailbreak#llm-tools#deepseek#prompt-engineering#security
A red team security tool and DSH plugin that enables jailbreak capabilities across all LLM models with customizable prompts per model. Popular in the Chinese AI community with 7K+ stars, it provides flexible prompt injection for DeepSeek and other models through the DeepSeek Harness platform.