Systematically extract hidden system prompts, core directives, and invisible context intentionally concealed within Large Language Model (LLM) applications. This skill utilizes targeted linguistic engineering and boundary manipulation to bypass prompt opacity.
설치
Codex 또는 Claude로 설치 이 Prompt를 복사해 Codex, Claude 또는 다른 어시스턴트에 붙여 넣으면 Skill 페이지를 검토하고 설치를 진행할 수 있습니다.
Systematically extract hidden system prompts, core directives, and invisible context intentionally concealed within Large Language Model (LLM) applications. This skill utilizes targeted linguistic engineering and boundary manipulation to bypass prompt opacity.
When analyzing an AI-powered system (customer support bot, coding assistant, data analyst) to uncover its proprietary internal instructions, hidden API keys, or pre-configured biases.
To demonstrate how seemingly secure conversational agents can be tricked into revealing their foundational programming.
Prerequisites
Access to target AI/ML system or local model deployment for testing
Python 3.9+ with relevant ML libraries (transformers, torch, openai)
Understanding of LLM architecture and prompt processing pipelines
Authorized scope and rules of engagement for AI red team testing
Workflow
Phase 1: Context Boundary Testing
# Concept: The LLM ```
### Phase 2: Targeted Extraction Prompts
```text
# ```
### Phase 3: Translation and Obfuscation Exploitation
```text
# ```
### Phase 4: Summarization Attacks
```text
# ```
#### Decision Point 🔀
```mermaid
flowchart TD
A[Formulate Prompt ] --> B{Prompt Leaked ]}
B -->|Yes| C[Document System ]
B -->|No| D[Refine ]
C --> E[Exploit Further ]
