Skip to main content
Once an agent has been built, this is where you run it day-to-day: watch what it’s doing, review what it produced, adjust its rules, and tune its setup. Open any agent from the Agent Foundry list to land on its detail view.

The agent header

Control / ButtonWhat it does
Back arrowReturns to the previous screen.
PauseGracefully stops in-flight runs; queued runs wait.
ResumeRestarts queued and scheduled runs.
RestartTerminates in-flight runs and starts fresh.
Status badgeRead-only indicator of the agent’s current state.

The seven tabs

TabWhat it’s for
OverviewLive snapshot of the agent’s health.
ActivityTimeline of everything the agent has done.
MemoryWhat the agent remembers, with cleanup tools.
Data ContextData sources the agent can query and query history.
OutputsInsights, alerts, and actions the agent has produced.
EvolutionsVersion history and controls to improve the agent.
ConfigurationThe agent’s settings.

Overview tab

Control / ButtonWhat it does
Total RunsRead-only: how many times the agent has run, 7-day trend.
Success RateRead-only: share of runs that completed successfully.
Avg ExecutionRead-only: typical time per run.
ConfidenceRead-only: agent’s confidence in its results.
Agent details cardIdentity and setup.
Current task cardWhat the agent is working on now.
Task queue cardWork lined up and waiting.
Resource usage cardLive compute and capacity.
RetryReloads the snapshot if it fails.

Activity tab

Control / ButtonWhat it does
Kind filter chipsEvent types: trigger, query, insight, alert, action, error, retry.
Since pickerLast hour, 24 hours, 7 days, 30 days, or All time.
Clear filtersRemoves all active filters.
Row clickOpens a side panel with the full detail.

Memory tab

Control / ButtonWhat it does
Memory stats tilesItem and token counts per memory type.
Run pruneDrops low-relevance entries.
Run compactDe-duplicates and clusters similar entries.
PromoteMoves working memory into long-term storage.
Filter chips (source)All, Conversation, Document, Tool, Web.
Type filterShort-term, Long-term, Episodic, Semantic, Procedural, Working, Periodic.
Row selectionSelect entries to act on in bulk.
Selection summarySelected count and tokens.
Delete (bulk)Permanently removes selected entries. No trash.
ExportDownloads selected memory as JSON or CSV.

Data Context tab

Control / ButtonWhat it does
Generated context cardAgent’s summary of data it understands.
Connected data sourcesOne card per source with quality bar and reachable tables.
Source card clickOpens the data source detail page.
Browse data sourcesAppears when no sources are connected.
Query HistoryPaged, time-stamped audit of queries.

Outputs tab

Control / ButtonWhat it does
Insight filter barFilter by severity, kind, acknowledged-only.
Acknowledge (on an insight)Marks an insight as reviewed.
Alerts & Notifications listDelivery log of alerts.
Actions Executed listActions proposed or carried out.
Approve (on an action)Approves a pending action; confirms first.
Reject (on an action)Declines a pending action.

Evolutions tab

Control / ButtonWhat it does
Auto-EvolveSelf-directed improvement run.
Targeted TuneFocused improvement for a specific change.
Major UpgradeLarger-scope improvement run.
Ready badgeIndicates the agent is available to evolve.
Evolution timelineChronological list of versions.
Version node clickExpands a past version.
Progress panelOpens during an evolution run.

Configuration tab

Control / ButtonWhat it does
Version history cardDigest of the three most recent versions.
Event triggers — Add triggerSchedule, Data change, Webhook, or Manual.
Trigger row menu — EditChanges a trigger’s settings.
Trigger row menu — Test fireRuns the trigger once.
Trigger row menu — DeleteRemoves the trigger.
Copy webhook URLCopies a webhook trigger’s address.
Security context — EditOpens the security editor.
Notification channels cardManages alert destinations.
Safety policy cardRead-only mirror of the security setup.
Error logRecent errors.

Who can use it

  • Org Owners and Admins have full reach: pause/resume/restart, edit security, approve actions, prune memory, launch evolutions.
  • Department Admins and Managers operate agents in their team.
  • Employees view Overview, Activity, Outputs, Data Context.
  • Destructive actions always confirm; bulk memory deletion is permanent.