Open database of system prompts extracted from frontier LLMs using JustAsk
Reverse-engineered system prompts from Claude (claude.ai) and Claude Code CLI, validated against Anthropic's officially published prompts.
A collection of extracted system prompts/operational instructions
Source code of the paper "Why Are My Prompts Leaked? Unraveling Prompt Extraction Threats in Customized Large Language Models"
AI 中转站 API 验真取证工具：模型验真 · 系统提示词提取 · 上下游供应链溯源 | LLM relay API forensics: verify models, extract system prompts, trace upstream supply chains
Official implementation of "ProxyPrompt: Securing System Prompts against Prompt Extraction Attacks"
Explore and compare 45+ system prompts across models with a live gallery, research paper, and open dataset for prompt analysis
To associate your repository with the prompt-extraction topic, visit your repo's landing page and select "manage topics."