Viewing Worker Details

The worker detail page shows the worker's performance metrics, latest findings, and configuration. See My Workforce to learn about the My Workforce page.

Breadcrumb

A breadcrumb at the top of the page shows the path to the worker, for example AI Workforce / Vulnerability Management / Exposures Triage. Select any part of the breadcrumb to navigate back.

Worker Header

The worker header shows the following:

The worker name and a category tag.
A star icon. To follow the worker, select the star (Follow). To unfollow the worker, select the filled star (Unfollow). See Following Workers.
The worker's category and schedule as a subtitle, for example "Vulnerability Management · After every discovery cycle".
A description of what the worker checks.
Edit — opens the worker's configuration for editing.
Run now — runs the worker immediately. Select the dropdown arrow next to Run now for additional run options.
On the right side: the time and type of the last run, for example "Last run 2 hours ago · scheduled", and Open conversation — opens the worker's AI chat conversation.
A status control in the top-right corner showing Enabled, Paused, or Enabled Read-Only. Select it to change the state. Disabled is only a badge on a worker that is missing a requirement, not a status value.

Headline Metric

Below the worker header, the page shows the worker's headline metric and how it changed since the last run. For example:

"Headline: 120 — open findings clearing the ticketing policy (of 4,500 open, non-excepted findings) · No change since last run"
"Headline: 0 — checks stale or never run · ↓ 1 fewer since last run"

The change indicator uses an arrow and a description, for example "↓ 1 fewer since last run" or "↑ 3 more since last run". When no change occurred, the indicator shows No change since last run.

Measured Runs Indicator

The Last N measured runs indicator below the headline shows a small trend chart for the most recent runs of the worker. Hover over the indicator to see details.

Performance Metrics

The detail page shows four summary metrics for the worker:

Runs — the total number of your runs of this worker.
Notifications — the total number of notifications from this worker, with the number unread, for example "2 unread".
Avg. run time — the average run duration over your last runs.
Actions applied — the number of actions applied by the latest run. The sub-label shows the scope, for example "by the latest run" or "previewed for your approval".

Ask AI

Select Ask AI in the top-right of the detail page to open the AI Companion with a question about the worker. For example:

Explain the Remediation Tracking worker on its AI Workforce page — what does it do, how did its recent runs go, and what should I do about what it found?

Overview Tab

The Overview tab shows the worker's latest findings and its configuration. It includes the following sections.

The Cadence, Standing instructions, and Approved Actions sections each have an Edit icon (pencil) in the section header. To edit one section, select its Edit icon. The edit form opens at that section. Edit in the worker header opens the same form at the top. To edit these sections, your role needs the Configure AI system settings permission. After you make your changes, select Save Changes.

Latest Briefing

The Latest briefing section shows the full output from the worker's most recent run. The section heading shows the time and source of the run, for example "Today at 3:43 AM · AI Workforce".

A briefing with findings has the same parts, in this order:

The head — The headline sentence, the Headline metric line, the Severity line, and what changed since the last run. A line separates the head from the other parts.
Summary — Three to five short bullets. They tell what the worker checked, what it found, what it means, what changed since the last run, and what the run did.
Findings — Up to five numbered findings, worst first. Each finding has a title, one or two sentences that explain it, and its evidence.
Options explored — What already exists to act on the result. This part appears only when the run looked for options.
Actions this run — What the run did with the worker's approved actions, in which mode, and the actions it didn't do. This part appears only when the worker has approved actions.
Queries used — The queries that the run counted with.
A last line that tells you how to tune the worker.

The briefing also ends with follow-up chips, for example Recommended actions that open a related worker or conversation. The last chip is Refine this worker.

The AI Workforce Manager briefing has its own parts after Summary: Run next, Stale or never run, Needs a clean run, and Quiet. It also shows Actions this run when the AI Workforce Manager has approved actions. See AI Workforce Manager.

Short answers keep their short form. An all-clear result, an empty window, and a direct question to a worker have no sections.

Briefings that a worker produced before 10.0.0 keep their earlier layout.

When the worker has never run from your own runs, the section shows "No completed briefing from your own runs yet." For the AI Workforce, it shows "No completed briefing yet — the AI Workforce's first run lands here."

General Settings

The General Settings section shows the worker's configuration details:

Worker Description — a description of what the worker checks and does.
Category — the department the worker belongs to.
Entry phrase — the phrase that runs the worker by hand when entered in the AI Companion, the same as Run now.

Cadence

The Cadence section shows the worker's schedule as a table with the following columns:

Schedule — the schedule name and run frequency, for example "After every discovery cycle".
Next run — when the worker is next scheduled to run. A dash (—) appears when the next run time isn't determined yet. When the worker status is Paused, the column shows "Paused — not running on schedule".
Last run — the date, time, and status of the most recent run, for example "2026-10-15 03:41 (Completed)".
State — the current state of the schedule, for example Active or Paused. When the worker status is Paused, every row shows Paused. Hover over the row to see "This worker is Paused, so it skips its scheduled runs. Set it to Enabled to run on schedule again."
Created by — who or what created the schedule, for example "AI Workforce" or your name.

To open the full list of scheduled workers, select Open Scheduled Workers.

To manage the cadence, use the following controls in the edit form: Change cadence, Restore default cadence, and Delete schedule. Save Changes saves a changed cadence with the other settings. See Changing AI Workforce Cadence.

Standing instructions

The Standing instructions section shows any persistent instructions that apply every time the worker runs, regardless of what the worker finds.

When no standing instructions are set, the section shows "No standing instructions yet." Standing instructions apply only in your data scope and take up to 4,000 characters.

Use standing instructions to change what the worker counts, and to add to what it reviews. For example, "Always add a review of the GenAI SaaS apps" adds that review to every briefing.

Approved Actions

To change settings on this page, you need one of the following permissions:

Configure AI system settings — changes status, Approved Actions, Max writes per run, and Standing instructions.
Manage the AI Workforce (with global data scope) — changes the AI Workforce cadence.

The Approved Actions section lists every action the worker can take and its current mode. Each action row shows:

The action name, for example Create Case, Create Ticket, Run Enforcement, or Run Workflow.
Three mode options: Human in the loop, Live, and Off. The active mode is highlighted.
- Human in the loop — the worker previews the action and waits for your approval before applying it. This is the recommended starting mode for most actions.
- Live — the worker applies the action automatically on every run without stopping for approval.
- Off — the worker doesn't take this action.
Additional details below the mode options, for example the connection or enforcement the action uses, or a note such as "No Enforcement Set bound."

Each action row also shows a Recommended starting mode badge and an Apply recommended configuration option. Every action starts in its recommended mode, and a recommended mode is never Live. The two actions that run an Enforcement Set or a workflow, Run Enforcement and Run Workflow, start at Off.

During a run, the worker performs its approved actions, including runs on a schedule and runs of the AI Workforce. The briefing lists the result of every approved action under Actions this run. An action that the worker didn't call is listed as not attempted, with the reason.

The section also shows Max writes per run — the maximum number of write actions the worker applies in a single run. The value is between 1 and 100, and the default is 10.

To change the mode of an action, select Edit, change the mode, and then select Save Changes. Changing an action to Live first shows a Switch to Live? confirmation dialog.

Running an Enforcement Set on a Run's Findings

Use Run on to run a bound Enforcement Set on the assets that the worker found in a run, instead of on the set's own trigger query. To change Run on, your role needs the Configure AI system settings permission.

In the Approved Actions section, select the Edit icon.
For Run Enforcement, select Live or Human in the loop, and then select the Enforcement Set.
Under the set, in Run on, select This run's findings.
Select Save Changes.
When the action is Live, the Run on this run's findings? dialog opens. To save, select Run on this run's findings.

Run on has the following options:

Its own trigger query (default) — The set runs on the assets of its saved trigger query.
This run's findings — The set runs on the assets that the worker's queries found in that run.

Note the following:

Run on appears only for Run Enforcement, and only after you select a set. Run Workflow has no Run on option.
With This run's findings, the set runs on at most 100 assets. When the findings match more than 100 assets, the set doesn't run, and the run report shows the count.
When the findings match no assets, the set doesn't run. The set doesn't use its trigger query instead.
Only Enforcement Sets on Devices or Users can run on a run's findings.
The Approved Actions section shows your choice under the set name, for example "Runs on: This run's findings". The line appears when the mode isn't Off.

For example, you bind the Enforcement Set "Isolate exposed hosts" to Run Enforcement in Live mode, with Run on set to This run's findings. The next run finds 12 exposed devices with no EDR record. The set runs on these 12 devices only.

Test Runs

Test Run isn't an approved action, and no Axonius worker can have it. A worker never starts a test run by itself. Some briefings offer a follow-up chip that starts a test run, for example Test-run the fix in an Automation Check briefing and Test-run the never-ran set in an Enforcement Dead Weight briefing. When you select the chip, the test run follows your Tool Permissions settings. By default, an approval card appears first.

Activity Log Tab

The Activity Log tab shows a chronological record of the worker's activity. The columns are Run date, Status, Duration, Owner, and Schedule activity. Select Show more runs to load older entries. You can investigate the details of a run in the AI Companion.

To start a conversation about a run:

In the Activity Log, locate the run you want to investigate.
In the row of the run, click Open conversation. The AI Companion opens in a side panel. The run is read only.
In the AI Companion, click Clone conversation to create a copy of the conversation that you can continue.

Notifications Tab

The Notifications tab shows all notifications this worker generated, with the most recent first. The tab heading shows the number of unread notifications, for example "Notifications 1".

Select a notification to open its details and act on any proposed changes. For more information, see Viewing AI Workforce Notifications.