Skip to content
Agentflow

Production Ops

production-ops owns:

How do we know production is healthy, and how do we recover when it is not?

Use when

Section titled “Use when”
  • a service is entering production
  • runtime health signals change materially
  • monitoring or alerting needs design
  • incident response is active
  • restore/recovery behavior needs proof
  • capacity or operational toil is the problem

Don’t use when

Section titled “Don’t use when”

Do not run an operations review for every code change.

Model

Section titled “Model”
runtime
  ↓
health evidence
  ↓
actionable event
  ↓
safe response
  ↓
recovery
  ↓
recovery verification

Try it

Section titled “Try it”
Use $production-ops to define health signals and recovery
for the booking service after the new database dependency.

Canonical runtime reference

Section titled “Canonical runtime reference”

Read the complete production-ops instructions.