signalisys
Readiness glossary

Trust & Governance

Your Einstein Trust Layer posture — PII masking, audit trails, prompt logging and toxicity controls that let agents act safely.

Why this score moves

  • Adoption of Einstein Trust Layer features in production
  • Whether evals run continuously, not only pre-launch
  • Maturity of incident response specifically for agents

Reading the four bands

EMERGING

Little to no logging of agent actions; no clear owner for safety, drift or incident response.

DEVELOPING

Logging exists but isn't analysed; policies sit in slides, not in production controls.

ESTABLISHED

Trust Layer controls live: PII masking, prompt + tool tracing, named owners and reviewed incident playbooks.

OPTIMISED

Continuous evals, drift detection and policy-as-code with measured time-to-detect / time-to-mitigate.

What good looks like

  • PII masking and toxicity filtering enforced on every prompt
  • End-to-end traces: prompt, action calls, data accessed, outcome
  • Named owners for safety, privacy and model drift
  • Measured detect / mitigate times for agent incidents
More dimensions
Data Cloud
Data Cloud Readiness
Agentforce
Agentforce Fit
Flow
Flow & Automation Depth
Core Cloud
Core Cloud Foundation
HITL
Human-in-the-Loop Design
Exec
Executive Alignment