L1Xu4n/Awesome-ChatGPT-prompts-ZH_CN

# Technical Summary A curated Chinese-language prompt engineering repository that documents jailbreak techniques and role-play templates for ChatGPT and Bing Chat, leveraging multi-step prompt injection chains to circumvent content policy guardrails. The collection includes specialized "wrapper" prompts that disguise restricted requests as functional testing or translation tasks, along with character-specific templates designed to establish persistent behavioral personas through iterative reinforcement across conversation turns. The project actively tracks model-specific vulnerabilities across Claude (v1/v2), Bing Chat, and GPT variants with dated updates reflecting real-time changes to model safety implementations and corresponding circumvention methods.

3,138 stars. No commits in the last 6 months.