Cencurity
Security gateway that proxies LLM and agent traffic to detect and redact secrets, PII and risky output for dev teams.
What it does
Cencurity is a security gateway that sits between AI coding agents and LLM providers, inspecting every request and response. It detects and masks secrets and sensitive data, blocks risky output, and logs interactions for audit. It works with any LLM or IDE without code changes.
Core features
Central dashboard of agent calls, latency and policy hits
Real-time detection and blocking of secrets and PII
LLM proxy with automatic data redaction
End-to-end log analysis and request tracing
Webhook alerts to Slack, Jira and more
Dry-run mode to test policies before enforcing
Best for
→Preventing prompt and data leakage in AI coding workflows
→Governing LLM use across teams and IDEs
→Producing audit-ready compliance evidence