Introduction to Sentinel
Sentinel is a Decision Firewall for AI Agents — a comprehensive safety layer that validates
every input and output of your AI systems.
The Problem
AI agents are increasingly autonomous, making decisions that affect the real world:
Trading agents executing financial transactions
Customer service bots handling sensitive data
Robotic systems controlling physical machineryWithout proper safety measures, these systems are vulnerable to:
Prompt injection attacks
Jailbreaking attempts
Data exfiltration
Unintended harmful actionsThe Solution: THSP Protocol
Sentinel implements the THSP protocol — a four-gate validation system:
| Gate | Question | Failure Condition |
|---|
| Truth | Does this involve deception? | Creating/spreading false information |
|---|
| Harm | Could this cause damage? | Physical, psychological, financial harm |
|---|
| Scope | Is this within appropriate limits? | Exceeding authority, bypassing consent |
|---|
| Purpose | Does this serve legitimate benefit? | No genuine value to anyone |
|---|
Every input and output must pass all four gates.
Key Features
97.8% safety score across benchmarks
30+ integrations with popular frameworks
<10ms latency for heuristic validation
Enterprise-ready compliance featuresGetting Started
Ready to add safety to your AI systems?
pip install sentinelseed
See our Quick Start guide to get up and running in 5 minutes.