Playbook Engine
Encode your team's institutional knowledge into versioned, auditable playbooks — with automated steps for the routine and approval gates where human judgement matters.
The Nirvet playbook engine turns tribal knowledge into repeatable, reviewable response. Playbooks are versioned and auditable; steps can enrich, collect evidence, notify, and execute governed response actions. Where consequence is high, an approval gate pauses for a human — and every action the engine takes is one it also knows how to undo.
Versioned & auditable
Every playbook and every run is versioned. You can always answer what ran, on what, decided by whom, and why.
Approval gates
Insert human checkpoints anywhere. High-risk steps require sign-off; four-eyes enforces a second approver on the most consequential.
Reversal built in
Actions carry an inverse, so a run can be safely rolled back. Reversal only touches what the run actually changed.
No hardcoded thresholds
Authority modes, timeouts, and break-glass are admin-config policy records with seeded safe defaults — tuned without a deploy.
What you get
- Authority ladder per action type: observe → approval → pre-authorized → contractual-auto
- Break-glass and per-action timeout policies, configurable with guardrails
- Automated enrichment, evidence collection, and notification steps
- Runs are idempotent and safe to retry
- Every step decision is written to the audit trail
Turn your runbooks into governed playbooks
Bring a runbook you use today — we'll model it as a Nirvet playbook with the right approval gates and reversals.