Brand Voice Guardian employee resolution
Resolution inspector for CLI, model, prompt and skill set by role, phase and severity.
Role
Brand Voice Guardian
brand_voice_guardian
Final model
fake-brand_voice_guardian
Fake Provider, enabled
Prompt
v1.0.0
editorial/brand_voice_guardian@1.0.0
Cost
~$0.18 / run
read-model estimate
global / global
Model resolution
per-severity override
no matching override in read model
per-stage override
no matching override in read model
role default: brand_voice_guardian
no matching override in read model
factory default: editorial/brand_voice_guardian@1.0.0
chosen: fake-brand_voice_guardian
api
Invocation state
Fake Provider
invocation enabled
passed
deterministic suite >= 90%
local shape only
Latest deterministic read model passed prompt, skill, and output-shape checks.
Constraints
Capability Card and matrix checks before invocation.
allowed providers: fake, claude_code, codex_cli, opencode, gemini_genai
preflight input
evidence pack required before invocation
preflight input
phase: 3
preflight input
ownsText: false
preflight input
allowedSkillKeys: editorial/brand-voice, projects/{slug}/brand-voice, projects/{slug}/forbidden-claims, editorial/issue-writing
preflight input
phase: 3
preflight input
deterministic: true
preflight input
Skills
Materialized skill set for this role.
editorial/brand-voice v1.0.0
loaded in prompt assembly
editorial/issue-writing v1.0.0
loaded in prompt assembly
Audit feedback
What the UI can state without mutation support.
matrix scope: global
read model
provider descriptor present
read model
invocation enabled by runtime
read model
changes require backend command route
read model
Prompt and output contract
Brand Voice Guardian operating prompt
Execute the brand_voice_guardian responsibility in the editorial runtime.
seed:editorial/brand_voice_guardian@1.0.0
editorial/brand_voice_guardian@1.0.0
editorial/brand_voice_guardian/output@1.0.0
Capability Card contract