Nova Notes is the in-platform knowledge base. Runbooks, postmortem templates, on-call notes, "how we deploy" pages. Every note is linkable from incidents, services, and decision bundles. The note about pgbouncer recovery shows up where the on-call engineer needs it: on the war-room page for a pgbouncer incident.
Every note attaches to one or more services, one or more incident classes, or both. When a payments incident opens, the war-room page shows all payments-scoped notes. When a "pgbouncer pool exhaustion" alert fires, the note tagged for that class is one click away. Notes that nobody references for a long time get flagged as candidates for archive.
Notes support embedded snippets that are more than text. NovaQL queries can be run from a note and the results inline. Bash blocks can be one-click copied. Approval-gated snippets ("run this on prod") link to Approval Queue rather than running directly. The note becomes a runnable runbook, not a stale doc.
Edits create new versions. Each version has an author, a timestamp, and an optional reviewer. Tier-0 service notes can require a reviewer before publishing. Old versions are recoverable; the version history is visible alongside the note. Notes are documents you can audit, not just text files.
Tenants can opt in to a shared library of community-authored runbooks for common technologies (Postgres, Redis, Kubernetes). Sharing is opt-in and goes only one direction: you read from the library, you do not contribute unless you explicitly publish. Library entries are read-only on your side; your customizations live in tenant-private notes.
Subscribe to Nova AI Ops on YouTube for demos, tutorials, and feature deep-dives.
Notes are useful when the engineer reads them. Linking them to incidents and services puts them where reading happens.