Observe
What the system records about itself while it works for you, and what you can do with it.
Everywhere else in this documentation, the fleet is the thing doing the work. This section is about the fleet as the thing being measured — what it notices about itself while it runs, what it is allowed to write down, and which of those things become decisions you make.
For an operator or an engineering lead who wants to know what the system is reporting about its own condition, and on what evidence.
After reading it you can say what makes a defect admissible here, what the counts on a queue are derived from, and which acts of repair are yours rather than the system's.
The pages of this section
- The defects your agents measured — Your agents file defects about the system they work in, each one carrying the command that measured it, and you decide what gets repaired.
Two neighbouring sections answer questions that sound like these and are not: Troubleshooting is what a refusal means when you are holding one, and Audit and compliance is what the record proves to somebody who was not there.
Worked examples
The commands, in the order the money actually moves: price it, check capacity, compose it, then read what it cost — and two refusals that cost nothing.
The defects your agents measured
Defects about the system itself are written down as files carrying the measurement that found them, checked against a fixed shape, and counted by class — and repair stays a human decision.