Skill Wiki
Permissions Human Readable
Permission and role labels in UI must describe what the user can do in plain language, not surface technical flag names, database identifiers, or internal role codes.
$ prime install @community/rule-permissions-human-readableProjection
Always in _index.xml · the agent never has to ask for this.
PermissionsHumanReadable [rule] v1.0.0
Permission and role labels in UI must describe what the user can do in plain language, not surface technical flag names, database identifiers, or internal role codes.
Loaded when retrieval picks the atom as adjacent / supporting.
PermissionsHumanReadable [rule] v1.0.0
Permission and role labels in UI must describe what the user can do in plain language, not surface technical flag names, database identifiers, or internal role codes.
Severity
medium
Loaded when retrieval picks the atom as a focal / direct hit.
PermissionsHumanReadable [rule] v1.0.0
Permission and role labels in UI must describe what the user can do in plain language, not surface technical flag names, database identifiers, or internal role codes.
Severity
medium
Examples
- Correct: 'Admin — can manage billing, invite members, delete projects'
- Correct: 'Viewer — can see all content but not make changes'
- Violation: Role dropdown showing 'admin', 'editor', 'viewer' with no description
- Violation: 'ROLE_MANAGER_L2' visible in permission settings UI
- Correct: permission request modals: 'Allow location access to show nearby results' not 'geolocation'
Severity
medium
Source
prime-system/examples/frontend-design/primes/compiled/@community/rule-permissions-human-readable/atom.yaml