AI-Assisted Configuration: From Point Lists to Operations
For first-time use, ask the Agent to start with aems-configure, which includes all API contracts and domain-specific guides. Edge-EMS Builder points, Lua scripts, and strategies can be configured through APIs. AI Agents working with Skills can assist with repetitive configuration, queries, analysis, and result verification. Engineers remain responsible for business objectives, strategy review, and site validation.
Choose Skills by task
Section titled “Choose Skills by task”| Stage | Work AI can assist with | Skill |
|---|---|---|
| Acquisition preparation | Convert point lists into Excel files importable by CommBridge | point-table-to-commbridge-excel |
| Data modeling | Create aggregation points, time schedules, and device templates | aems-collection-config, aems-add-aggregation, aems-time-schedule, aems-data-groups |
| Device expansion | Configure the three UDE template types and dynamic rules, run trials, and track materialization | aems-ude, aems-ude-placeholders |
| Strategy orchestration | Write Lua and configure triggers, conditions, actions, and post-processing | aems-lua-script, aems-add-trigger-plan |
| Alarms and communication | Configure alarm observations, MQTT, and real-time reads/writes | aems-alarm-lifecycle, aems-add-mqtt, aems-device-sharing, aems-write-data |
| Data analysis | Query trends, export details, and estimate future changes | aems-history-chart, aems-history-export, aems-forecast |
| Ongoing maintenance | Configure periodic archiving, assess capacity, and verify runtime configuration | aems-history-archive, aems-runtime-config, aems-backup-restore |
See all Skills and installation instructions. Point-list conversion only generates an import file; whether it is applied to a device depends on the actual delivery workflow. Forecasts are produced by the Agent’s analysis of historical data, not an Edge-EMS Builder built-in forecasting service.
Use aems-account-management and aems-display-config for account roles and display configuration respectively. The comprehensive maintenance entry point is aems-system-maintenance.
Recommended collaboration workflow
Section titled “Recommended collaboration workflow”- Describe the objective, device scope, required points, and constraints.
- The Agent queries actual configuration and produces parameters and scripts that can be checked.
- Validate conditions, missing data, and exception handling in a test environment before applying them to the target scope.
- Read back configuration, points, and runtime records, and continuously track results.
For batch template updates, validate one instance before expanding the reconciliation scope. Alarm rules should handle normal, abnormal, and unknown states. For exports and archives, check disk space, retention periods, and completion status.
APIs and permissions
Section titled “APIs and permissions”Agents use the same account permissions, maintenance freeze restrictions, and resource limits as other clients. Provide credentials in the runtime environment, not in Skill packages. Device protection, hardware interlocks, and site validation remain the responsibility of the relevant systems and engineering workflows.
Learn more about the Lua engine, alarm lifecycle, and operations and maintenance.