Content & Brand Governance exit conditions
Candidate exit conditions for the Content & Brand Governance type. See the catalog overview for how to adopt these, and Content governance for how the deterministic and holistic checks divide.
This type covers the language and tone of generated prose. It is the set a company most often replaces with its own version, supplied through the governance repository and governance-policy-url. Documentation completeness and structure stay in the Documentation & Training type.
|
Two ids on this page ship in the default library, both off by default (opt in per project or promote them in a policy): |
Language (deterministic, introspective Vale auditor)
| Exit condition | What it checks |
|---|---|
|
Prohibition, off by default. The prose passes the content class’s Vale configuration with a zero exit; a non-zero exit is red. |
|
No phrase on the banned list is present (Vale |
|
Mandated substitutions are honored, for example "use" not "utilize" (Vale |
|
Reading grade stays within the configured bound (Vale |
|
Product and feature terms match the approved terminology, for example "Akka Specify Plugin". |
|
The Vale package matches the version the governance policy pins. |
Tone (holistic, delegated to a judge)
| Exit condition | What it checks |
|---|---|
|
Completion, delegated, off by default. The prose passes the decomposed adversarial rubric, recorded by an attestation keyed to a content-plus-rubric signature. |
|
The tone judge runs a model from a different family than the one that generated the text. |
|
The recorded attestation covers the current rubric, and re-stales when the rubric changes. |
|
Case-study and marketing prose states verifiable facts rather than unsupported claims. |
Per-content-class scoping
| Exit condition | What it checks |
|---|---|
|
Every governed content path resolves to a content class with an attached exit-condition set. |
|
A separate marketing repository governed by policy has its content class enforced, not only defined. |