Skip to content

Department · Operations & Infrastructure · schema_pack: 6 types

The runbook should know the infrastructure changed.

Runbooks and postmortems checked against the real system, drift flagged before the shift, and context in the incident copilot.

Scenario

How context circulates when ValorBrain is in the flow.

  1. 01

    Checking the payment gateway's failover runbook against the real system finds the documented hostname different from the current ingress in k8s.

  2. 02

    The runbook gets marked as divergent, with the owner and the date of the last check.

  3. 03

    The divergence becomes a review for sre/platform, not a surprise in the middle of an incident.

  4. 04

    A similar incident retrieves the earlier postmortems, with what worked and the source of each decision.

  5. 05

    The incident copilot starts from that history instead of re-discovering the cause.

  6. 06

    What was learned last time shows up in the first answer this time.

  7. 07

    Whoever comes on call receives in the session bootstrap what changed since the last shift: drift under review, a pending handoff, a new decision.

  8. 08

    The handoff between shifts and agents is routed with state, not summarized from memory.

  9. 09

    The shift starts from what changed, not from what the runbook forgot.

Schema pack

Unit types that give structure to what's worth remembering in this area. Every record carries source, owner, version and validity.

schema_pack
schema_pack:types:  - runbook  - postmortem  - sla  - dependency  - oncall  - infra_change

Next step

Less starting over. More continuity

Bring a real project. In one conversation we map where context gets lost between the people and agents on your team, and where to start.