Overview

AI testing agents extend the Zof Console with specialized validation capabilities, accessibility, performance, security, API contract, and custom enterprise workflows. Agents are catalog offerings enabled per organization and executed under policy through Agent Console or integrated pipelines.

This documentation covers agent concepts, fleet operations, cloud and endpoint execution models, and custom agent development. Use it when scaling from ad hoc Console runs to governed, fleet-based validation across staging, production, and restricted environments.

Agent Console is the operational surface for execution agents inside the Zof Console product. It is not the product name itself, always distinguish Zof Console (control plane) from Agent Console (execution fleet).

Who should read this

QA engineers, SREs, platform teams, and developers operating Zof Console and APIs.

Prerequisites

Organization membership with permission to view Automation and Agent Console areas
At least one project with reviewed test cases or suites ready for execution
For endpoint workloads: network and security approval from your infrastructure team

When to use this workflow

Onboarding new team members to Zof terminology and workflows
Authoring internal runbooks aligned with Console labels
Designing CI/CD or webhook integrations against documented behavior

Step-by-step procedure

Assess your execution model

Identify application types (web, API, desktop, hybrid) and target environments (staging, production, isolated networks).

Determine whether cloud agents, endpoint agents, or both are required for your compliance and connectivity constraints.

Document expected concurrency, data residency, and artifact retention requirements before fleet rollout.

Enable agents from the catalog

Open Automation → AI testing agents in the Zof Console.

Review agent capabilities, supported platforms, and policy requirements for each catalog entry.

Enable agents approved by your organization and assign ownership to QA, SRE, or platform teams.

Configure Agent Console fleet

Navigate to Automation → Agent Console for fleet overview and health metrics.

Register cloud agent pools or deploy endpoint agents according to your environment topology.

Apply labels and capability tags so orchestration routes jobs to appropriate executors under policy.

Connect applications and execution targets

Register applications in Agent Console with environment metadata, authentication context, and team ownership.

For desktop or VDI workloads, create endpoint applications linked to registered endpoint agents.

Validate connectivity with a smoke run before scheduling release-critical validation.

Execute and monitor at scale

Launch runs from Agent Console, Console Runs, schedules, or CI/CD integrations.

Monitor execution telemetry, queue depth, and agent heartbeat during peak release windows.

Review results in Runs and Test Health; escalate infrastructure failures through operational runbooks.

Govern and iterate

Align agent usage with roles, permissions, and audit requirements in Admin Center.

Review fleet utilization quarterly and right-size cloud pools or endpoint capacity.

Extend capabilities with custom agents when catalog offerings do not meet specialized needs.

Key concepts

Organization scope: All Zof Console and API operations are isolated to your authenticated tenant.
Governed execution: Agent output and remediation follow policy packs with human approval when configured.

Best practices

Treat agent labels as part of your routing contract, document label conventions in team runbooks.
Separate staging and production agent pools to prevent cross-environment data leakage.
Require human review of generated tests before routing them to production-facing agent pools.
Monitor agent heartbeat and queue latency as first-class SLOs during release cycles.
Use least-privilege roles for operators who manage fleet configuration versus those who only launch runs.

Common issues

Jobs remain queued with no matching agent: Label, capability, or environment metadata on the run does not match any online agent. Reconcile policy targeting in Agent Console and verify agent registration.
Endpoint agent shows offline after deployment: Network egress, certificate trust, or host firewall rules may block control-plane connectivity. Confirm outbound HTTPS access and restart the agent service after policy updates.
Cloud agent pool exhausted during peak load: Concurrent run demand exceeds configured pool capacity. Increase pool size, stagger schedules, or prioritize critical suites through governance policies.