Lab 04: Data Leakage and System Prompt Leakage
Secret directive exfiltration techniques (meta-prompt, format bypass), synthetic Canary Token deployment, and Output Redactor pipeline.
Key Lab Topics
- Canary Tokens
- Output Sanitization
- Instruction Anchoring
- Superficies de Exfiltración
Prerequisites
- Python 3.10+
- Regex / Canary Tools
Resources & Architecture Guide
This hands-on lab runs in a Jupyter Notebook (lab-04-fuga-datos.ipynb) and includes a comprehensive architectural guide: