Skip to content

AI Features

Breeze includes a built-in AI assistant powered by the Claude Agent SDK that can query your fleet, diagnose device issues, and take action on your behalf. The assistant runs as a managed agent with a query loop — it receives your question, plans a sequence of tool calls, executes them against live fleet data, and returns a synthesized answer. All tool executions are governed by a tiered approval system with full audit logging and per-tool rate limits.

Six capabilities extend and govern this system:

  • AI Risk Engine — governance dashboard for controlling and auditing AI-assisted operations
  • Fleet Orchestration Brain — command center for fleet-scale AI-driven management
  • AI Device Context Memory — persistent per-device memory that the AI carries across conversations
  • Helper AI Chat — end-user-facing AI assistant in the Breeze Helper tray app
  • AI Cost Tracking — per-session and per-organisation usage and cost monitoring
  • AI Input Sanitization — guardrails that validate and sanitize all AI inputs before execution

The AI Risk Engine categorises every AI tool into one of four tiers that control how it executes. Navigate to it via Monitoring → AI Risk Engine in the sidebar.

TierExecutionExamples
Tier 1Auto-execute (read-only, no approval)query_devices, analyze_metrics, get_security_posture, get_active_users, file_operations (list/read), disk_cleanup (preview), propose_action_plan, list_configuration_policies, get_effective_configuration, preview_configuration_change, get_configuration_policy, configuration_policy_compliance, manage_processes (list), get_service_monitoring_status, query_monitors, query_backups, get_backup_status, browse_snapshots, get_incident_timeline, generate_incident_report, list_playbooks, get_playbook_history, search_documentation
Tier 2Auto-execute + audit loggedmanage_alerts acknowledge/resolve actions, manage_services list action, set_device_context, resolve_device_context, apply_configuration_policy, remove_configuration_policy_assignment, manage_configuration_policy (activate/deactivate), create_incident, collect_evidence
Tier 3Requires human approval before executionexecute_command, run_script, disk_cleanup (execute), network_discovery, security_scan (quarantine/remove/restore), file_operations (write/delete/mkdir/rename), manage_configuration_policy (create/update/delete), manage_processes (kill), manage_monitors (create/update/delete), trigger_backup, restore_snapshot, execute_containment, execute_playbook
Tier 4Blocked — never executedCross-org operations

When the AI proposes a Tier 3 action, it enters a pending state and waits for human approval.

  1. Open Monitoring → AI Risk Engine.

  2. Click Approval History.

  3. Find the pending request and review the action details.

  4. Click Approve to allow execution, or Reject to cancel it.

Approved actions execute immediately. Rejected actions are logged and the AI is notified.

Each tool has a per-tool sliding window rate limit. Requests that exceed the limit are rejected and logged in the Rejection & Denial Log.

ToolLimitWindow
execute_command10 requests5 min
run_script5 requests5 min
disk_cleanup3 requests10 min
network_discovery2 requests10 min
security_scan3 requests10 min
file_operations20 requests5 min
manage_services10 requests5 min
analyze_disk_usage10 requests5 min
get_user_experience_metrics20 requests5 min
manage_configuration_policy20 requests5 min
manage_processes15 requests5 min
query_monitors30 requests5 min
manage_monitors10 requests5 min
get_service_monitoring_status30 requests5 min
query_backups20 requests5 min
get_backup_status20 requests5 min
browse_snapshots20 requests5 min
trigger_backup3 requests10 min
restore_snapshot3 requests10 min
create_incident5 requests5 min
execute_containment5 requests10 min
collect_evidence10 requests5 min
get_incident_timeline20 requests5 min
generate_incident_report10 requests5 min
list_playbooks20 requests5 min
execute_playbook3 requests10 min

The AI can propose multi-step action plans for complex operations that require human review before execution. When the AI invokes propose_action_plan, it creates a structured plan with pre-flight checks for each step. Plans operate in two modes:

ModeDescription
action_planA sequence of concrete steps to be executed after approval
hybrid_planA mix of informational analysis and executable steps

Action plans are subject to the same approval workflow as Tier 3 tools — the plan must be approved before any steps execute.

The Risk Engine dashboard provides five views, each filterable by time range (24 h / 7 d / 30 d):

SectionWhat it shows
Tier Overview MatrixTool counts per tier with colour-coded risk levels
Tool Execution AnalyticsExecution status breakdown, top tools, average duration
Approval HistoryPending, approved, and rejected Tier 3 requests
Rate Limit StatusPer-tool limit cards with current usage
Rejection & Denial LogFailed, rejected, and security-denied operations
MethodPathDescription
GET/ai/admin/tool-executionsTool execution analytics (?since=ISO&limit=1–200, default 100)
GET/ai/admin/security-eventsGuardrail audit trail (?since=ISO&limit=1–100&action=filter)

The Fleet Orchestration Brain is an AI command centre for fleet-scale operations. Open it via Fleet in the main navigation sidebar.

The page shows eight stat cards that aggregate live fleet data:

CardWhat it shows
PoliciesTotal policies, enforcing count, compliance %, non-compliant devices
DeploymentsActive, pending, completed, and failed deployment counts
PatchesPending approval, approved, installed; critical pending count
AlertsCritical, high, medium, and low alert counts
Device GroupsCount of static and dynamic groups
AutomationsConfigured automations with recent run history
Maintenance WindowsActive windows with suppression flags
ReportsAvailable report templates and schedules

When you are on the Fleet Orchestration page, the AI assistant gains access to eight fleet-level tools:

ToolWhat it does
manage_policiesList, evaluate, create, activate/deactivate, and remediate policies
manage_deploymentsCreate, start, pause, resume, and cancel deployments
manage_patchesScan, approve, decline, defer, bulk approve, and rollback patches
manage_groupsCreate static/dynamic groups and manage membership
manage_maintenance_windowsSchedule maintenance windows with timezone support
manage_automationsCreate and update automation rules and event triggers
manage_alert_rulesConfigure alerting templates per device or site
generate_reportGenerate inventory, compliance, performance, and executive summary reports

The page includes pre-populated AI chat buttons that open the AI sidebar with a domain-specific prompt:

ButtonPre-filled prompt
Check complianceShow me a compliance summary for all policies
Active deploymentsList all active deployments and their progress
Critical patchesWhat critical patches are pending approval?
Alert overviewGive me a summary of active alerts by severity
Maintenance windowsWhat maintenance windows are active right now?
Run automationsList all enabled automations and their recent run history
Device groupsShow me all device groups and their member counts
Generate reportGenerate an executive summary report for the fleet

The AI can remember device-specific facts across conversations. When you ask the AI about a device, it automatically loads that device’s context entries and incorporates them into its analysis — so it won’t re-alert on known quirks or forget about open follow-ups.

TypePurposeExample
issueKnown problems to track”Recurring BSOD on boot since Jan 2026”
quirkNormal but unusual behaviour”Slow startup is expected due to a legacy driver”
followupPending actions”Check disk health after replacement on 2026-03-01”
preferenceUser or device preferences”Maintenance window: Sundays 2 AM–4 AM only”

Context is managed through the AI assistant — there is no separate UI. Ask naturally:

  • “Remember that this device has a recurring BSOD issue.”
  • “Mark the disk check follow-up as resolved.”
  • “What do you know about HOSTNAME?”

Context entries can have an expiry date, which is useful for time-bound follow-ups. Expired entries are excluded from future queries but are not deleted.

The AI uses three tools internally to manage context:

ToolTierDescription
get_device_contextTier 1Load context entries for a device
set_device_contextTier 2Create a new context entry
resolve_device_contextTier 2Mark an existing entry as resolved

The Breeze Helper tray application includes an AI chat interface designed for end users (not just administrators). When enabled at the organisation level, end users can ask the Helper questions about their device, troubleshoot issues, and request common IT actions — all without opening a support ticket.

The Helper AI agent runs on the API server using a filtered tool set. Not all AI tools are available to the Helper — the helperToolFilter service restricts access to safe, device-scoped operations appropriate for end users.

CapabilityDescription
Device diagnostics”Why is my computer slow?” — analyzes CPU, memory, disk, and boot data
Vision troubleshootingCaptures a screenshot and uses AI vision to interpret error dialogs or UI issues
Self-service actionsCommon IT tasks like clearing temp files, restarting services, checking updates
Computer controlAI can interact with the desktop to perform guided actions (Tier 3, requires approval)

Helper chat is controlled per organisation. Administrators can enable or disable it via organisation settings. The toggle is delivered to agents through the heartbeat response — when disabled, the Helper hides its chat interface.


Every AI session tracks token usage and estimated cost. The aiCostTracker service records input tokens, output tokens, and total cost per session, per user, and per organisation.

Navigate to Settings → AI Usage to view:

MetricDescription
Total sessionsNumber of AI conversations
Total tokensCombined input + output tokens consumed
Estimated costDollar cost based on Claude API pricing
Per-user breakdownToken usage grouped by user
Per-session detailIndividual session cost and token counts

Cost data is stored in the aiSessions and aiMessages tables alongside the conversation history.


The aiInputSanitizer service validates all user inputs before they reach the AI agent. This layer prevents prompt injection, command injection, and other adversarial inputs from being passed to AI tool calls.

Sanitization includes:

  • Input length limits
  • Pattern detection for known injection techniques
  • Escaping of shell metacharacters in tool arguments
  • Validation against Zod schemas defined in aiToolSchemas.ts

All rejected inputs are logged in the security event trail accessible via the Risk Engine dashboard.


The AI assistant can query, manage, and operate backups across your fleet. These tools are available in any AI chat session.

ToolTierDescription
query_backups1List backup configurations, jobs, and policies
get_backup_status1Health summary for a device or organization (active configs, job counts, storage totals)
browse_snapshots1List available snapshots for a device with timestamps, sizes, and expiration
trigger_backup3Initiate an on-demand backup for a device (requires approval)
restore_snapshot3Restore data from a snapshot to a device (requires approval)

See Device Backup for configuration details.


The AI can create and manage security incidents, execute containment actions, and collect forensic evidence.

ToolTierDescription
create_incident2Create a new security incident with classification and severity
execute_containment3Run a containment action on a device (process kill, network isolation, account disable, USB block)
collect_evidence2Collect forensic evidence from a device (logs, processes, connections, screenshots)
get_incident_timeline1View the full timeline of an incident
generate_incident_report1Generate a structured report with action and evidence summaries

See Incident Response for the full workflow.


The dashboard includes a slide-out documentation panel that opens the correct docs page based on your current location in the app. Open it with Cmd+Shift+H (macOS) or Ctrl+Shift+H (Windows/Linux).

Page you’re onDocs page that opens
ScriptsScripting docs
PatchesPatch management docs
DevicesDevice management docs
Configuration PoliciesPolicy docs

The help panel coordinates with the AI chat sidebar — opening one closes the other. When the AI assistant references documentation in a chat response, the links open directly in the help panel instead of a new browser tab.

The AI assistant can search all documentation pages using the search_documentation tool (Tier 1, auto-execute). The search index is built at compile time and covers all pages in the docs site.

ToolTierDescription
search_documentation1Search documentation pages by keyword and return relevant sections with direct links

The AI can list and execute remediation playbooks as part of automated incident response.

ToolTierDescription
list_playbooks1List available playbooks, optionally filtered by category
execute_playbook3Execute a playbook on a device with runtime variables (requires approval)
get_playbook_history1View execution history for a playbook or device

See Playbooks for playbook definitions and step types.


Tier 3 action pending but never executing

Tier 3 actions require manual approval. Open Monitoring → AI Risk Engine → Approval History and approve or reject the pending request.

AI Risk Engine dashboard shows no data

The dashboard requires at least one AI tool execution to have occurred. Ask the AI assistant a question about your fleet to generate initial data.

Fleet Orchestration stat cards showing zeros

Some endpoints (deployments, reports) return empty results if no data exists yet. Cards populate independently — a zero on one card does not indicate a general problem. Partial endpoint failures are shown as warnings in the UI.

Context entries not appearing for a device

The AI loads context only for the specific device you ask about. Try: “What do you know about [hostname]?” to trigger explicit context loading.

set_device_context not working

set_device_context is Tier 2 (auto-execute + audit logged) and requires devices:write permission. Confirm your role includes write access to devices.