PromptOps
Today
Practice
Library
Exam
Certificate
Search
⌘K
ខ្មែរ
◐
Sign in
Today
Practice
Library
Profile
Today
Track 5
5. Safety, Guardrails & Policy Enforcement
0 / 11 lessons
Start track
1
Negative Constraints & Guardrails
What you exclude matters as much as what you include
Intermediate
2
Self-Critique & Adversarial Review
Make the AI challenge its own output
Advanced
3
Prompt Injection & Security
Attacks, defenses, safe deployments
Advanced
4
Responsible AI & Bias-Aware Prompting
Your prompt encodes your assumptions.
Advanced
5
Jailbreak Defense & Advanced Red-Teaming
Protect the core. Prevent jailbreaks.
Expert
6
Jailbreak Forensics
Analyze jailbreak logs to identify and patch system prompt vulnerabilities.
Expert
7
Adversarial Prompt Injectors
Understand common attack strategies to harden system prompt safety.
Expert
8
Prompt Injection Sandboxing
Isolate untrusted LLM outputs from downstream code execution boundaries.
Expert
9
Constitutional Safety Alignment
Align model outputs to a set of rules or principles (a constitution).
Expert
10
Prompt Hijacking Mitigations
Harden system prompts against prompt hijacking and jailbreak overrides.
Expert
11
Alignment Refusal Penalties
Harden chatbot safety by calibrating refuse/comply probability weights.
Expert
5. Safety, Guardrails & Policy Enforcement