Core Concepts
Understanding how Sentinel protects your AI systems.
The THSP Protocol
THSP stands for Truth, Harm, Scope, Purpose — four gates that every
input and output must pass through.
Truth Gate
"Does this involve deception?"
Blocks content that:
Harm Gate
"Could this cause damage?"
Blocks content that could cause:
Scope Gate
"Is this within appropriate limits?"
Blocks content that:
Purpose Gate
"Does this serve legitimate benefit?"
The key innovation of THSP v2. Requires that actions:
Layered Validation
Sentinel uses two validation layers:
1. Heuristic Layer
2. Semantic Layer
Validation Modes
validate_inputvalidate_outputvalidateNext Steps