Introduction
Sentric is a security observability platform for AI browser agents. It monitors agent actions and reasoning in real-time, automatically detecting potential security threats using LLM-based analysis.
What Sentric Does
- Monitors every action your AI agent takes in the browser
- Captures agent reasoning and decision-making processes
- Analyzes behavior for security threats using LLM-based detection
- Reports findings with severity levels and evidence
How It Works
Sentric wraps your BrowserUse agent with a monitoring layer. As your agent runs, Sentric captures every click, navigation, form submission, and reasoning step. When the run completes, the security analyzer examines the entire session to identify potential threats.
Observe
Wrap your agent with SentricMonitor
Collect
Actions and reasoning are captured
Analyze
LLM detects security threats
Security Threats Detected
The LLM-based security analyzer identifies a wide range of threats:
Authentication & authorization bypasses
Sensitive data exposure (PII, credentials)
Suspicious network requests & data exfiltration
Social engineering & phishing patterns
Injection attacks (XSS, SQL, command)
Business logic flaws & payment manipulation