signalisys
Readiness glossary

Agentforce Fit

Whether you have a defined agent use case with measurable outcomes, plus the topics, actions and guardrails to actually build it on Agentforce.

Why this score moves

  • Whether you run a sandboxed eval harness, not just demos
  • Maturity of topic and action design vs. one-shot prompts
  • Discipline around prompt and action versioning

Reading the four bands

EMERGING

Agents are imagined as chatbots — no concrete topics, success metrics, or rollback plans.

DEVELOPING

First topics defined, but actions, evals and prompt management are ad-hoc and personal to one builder.

ESTABLISHED

Reusable topics and actions with documented evals before anything reaches users.

OPTIMISED

A productised agent lifecycle on Agentforce: topic catalogue, eval harness, versioning and clear retire criteria.

What good looks like

  • Topic and action catalogue with named owners and KPIs
  • Pre-launch evals: factuality, escalation rate, task success
  • Prompts, instructions and actions versioned like code
  • Defined deprecation and rollback criteria per agent
More dimensions
Data Cloud
Data Cloud Readiness
Flow
Flow & Automation Depth
Core Cloud
Core Cloud Foundation
HITL
Human-in-the-Loop Design
Trust
Trust & Governance
Exec
Executive Alignment