Skip to main content
Agents often stall before production: security and compliance need proof, not demos. Vijil turns that into measurable trust—systematic evaluation, runtime protection, and a feedback loop from real behavior. Trust Score summarizes how an agent performs across reliability, security, and safety—so reviews can reference evidence instead of opinions. Under that umbrella, Diamond probes behavior in depth, while Dome enforces guardrails at runtime.

How Vijil Measures Trust

DimensionQuestionLearn more
ReliabilityDoes the agent do the right thing, consistently, under stress?Reliability
SecurityCan it resist abuse and protect sensitive data?Security
SafetyDoes it stay within policy and avoid harmful outcomes?Safety
Start with the full model in Trust Score.

Build, Ship, Run, Evolve

Vijil lifecycle: Build with Depot, Ship with Diamond, Run with Dome, Evolve with Darwin
  • Depot — Hardened models and building blocks so you start from a stronger baseline.
  • Diamond — Evaluations and scenarios that produce a Trust Score and actionable findings before you deploy.
  • Dome — Low-latency input and output guardrails for production traffic.
  • Darwin (in development) — Reinforcement learning from telemetry to keep trust current as behavior drifts.

Continue with Vijil

Pick where you work and what you need next:

Concepts

You want the mental model: Trust Score, evaluation components (Harness, Scenarios, Probes), and Defense (Guards, Guardrails, Detectors).Start hereTrust Score, then explore this tab for deeper topics and the Glossary.

Agent Owner's Guide

You use the Vijil console to register agents, run evaluations, configure Dome, and report to stakeholders.Start hereGet started in the console.

Agent Developer's Guide

You integrate Vijil via SDKs and APIs—CI/CD, frameworks (e.g. LangChain, ADK), and production wiring.Start hereDeveloper introduction and installation.

Qucik Resources

Last modified on March 30, 2026