Playbooks
Codify your best remediation steps and run them consistently every time.
Playbooks turn institutional knowledge into repeatable, auditable workflows — so every technician on the team responds to common incidents the same way, every time.
Structured Multi-Step Workflows
Each playbook defines an ordered sequence of steps across five types: diagnose, act, wait, verify, and rollback. This structure ensures remediation follows a complete, consistent pattern — capturing baseline state before acting, pausing for systems to stabilize, and confirming success before closing the loop.
Built-in Playbooks for Common Scenarios
Breeze ships with three ready-to-run playbooks covering disk cleanup, service restart with health verification, and memory pressure relief. Each built-in playbook is fully auditable, uses the same tools available to the AI assistant, and can be executed manually or triggered automatically in response to matching alerts.
Custom Playbooks for Your Procedures
Organizations can author custom playbooks scoped to their own environment. Each step supports template variables for device-specific values, making it straightforward to build a single playbook that works correctly across an entire fleet without hardcoding device identifiers or service names.
AI-Triggered Incident Response
The Breeze AI assistant can execute playbooks automatically as part of an alert response conversation. When the AI detects a matching condition, it runs the appropriate playbook, monitors each step, and updates the execution record in real time — turning routine incidents into hands-free resolutions.
Controlled Failure Handling
Each step defines what happens when it fails: stop execution immediately, continue to the next step, or trigger a rollback sequence that undoes prior changes in reverse order. Teams can tune failure behavior to match the risk profile of each operation rather than accepting a one-size-fits-all approach.
Complete Execution History
Every playbook run is recorded with per-step timing, tool inputs and outputs, pass/fail status, and who or what triggered the execution. Teams have a full audit trail of every remediation action taken — supporting both post-incident review and compliance reporting for managed clients.
Capabilities
Multi-Step Remediation Workflows
Ordered steps of type diagnose, act, wait, verify, and rollback encode standard operating procedures as repeatable runs.
Built-In and Custom Playbooks
Three built-in playbooks for disk cleanup, service restart, and memory relief; organizations can define custom workflows.
Three Failure Policies
Steps can stop execution, continue past failure, or trigger a rollback sequence depending on their onFailure setting.
Full Execution Audit Trail
Per-step timing, tool input/output, and status are recorded for every playbook run with query and filter support.