# Techniques for Extracting Hidden System Prompts

> Master advanced techniques for extracting hidden system prompts, including continuation attacks, delimiter manipulation, and encoding obfuscation.

Canonical: https://codelivly.com/rooms/techniques-for-extracting-hidden-system-prompts
Last updated: 2026-10-10

| | |
|---|---|
| Level | Intermediate |
| Category | Cybersecurity |
| Lessons | 4 |
| Estimated time | 30 min |
| XP | 100 |

This module explores the specific, actionable techniques attackers use to extract hidden system prompts from LLMs. You will learn how to leverage continuation attacks, formatting manipulation, and logical paradoxes to bypass RLHF safety filters and force the model to reveal its foundational instructions, alongside the architectural defenses required to stop them.

## Lessons

1. Direct Ask and Repetition-Based Extraction (30 min)
2. Completion and Continuation-Based Extraction (30 min)
3. Translation and Encoding-Based Extraction (30 min)
4. Guided Lab: Extracting a System Prompt from a Test Target (30 min)

## Related

- [Learning path: System Prompt and Data Extraction](https://codelivly.com/learning-paths/system-prompt-and-data-extraction)
- [All rooms](https://codelivly.com/rooms)