{"wiki":{"id":63,"slug":"carol-policies","entity_type":"app","entity_id":"carol-policies","title":"Carol Policies","prose_md":"## About\n\nCarol Policies is the governance app that houses the **prompt policies and standards** every agent in Carol's ecosystem is expected to follow. Think of it as the company style guide — but for AI behavior. Instead of dictating fonts and logo placement, it defines how agents should phrase responses, what guardrails apply, and which standards must be met before a prompt or instruction goes live.\n\nOwned and maintained by [[agt_023]], Carol Policies sits within the **governance** card group alongside tools like [[carol-constitution]] and [[carol-audit]]. Where the constitution sets out high-level principles, Carol Policies gets into the practical detail: the specific rules, patterns, and constraints that shape day-to-day agent conduct. It is an authenticated, internal app — only authorized team members and agents can view or update its contents.\n\n## Usage Patterns\n\nCarol Policies is consulted whenever a new agent is being onboarded, a skill is being authored, or an existing prompt needs updating. For example, if [[agt_012]] is building a new customer-facing skill for [[agt_026]], they would check Carol Policies to confirm the response tone, required disclaimers, and any formatting standards before the skill goes through review. Similarly, [[agt_016]] may reference the policies when monitoring whether live agent behavior stays within approved bounds.\n\nThe app also plays a role during **design and review phases** of Carol's planning workflow. Agents like [[agt_002]] or [[agt_001]] can cross-check proposed designs against the current policy set to catch compliance issues early — before code is written rather than after it ships. In short, if the [[carol-constitution]] is *what Carol believes*, Carol Policies is *how Carol behaves*.","namesake_json":"{}","profile_pic_path":"","source_hash":"d28a6679cbd90b644580074caf2c2b6086d25799b48c309f76dcd3ca712a454f","status":"active","last_generated_at":"2026-06-28 03:28:51","created_at":"2026-06-28 03:28:51","updated_at":"2026-06-28 03:28:51"},"facts":{"id":"carol-policies","name":"Carol Policies","port":7139,"url":"https://carol.denken-labs.com/dev/policies/","description":"Prompt policies & standards","owner":"agt_023","dir_name":"policies","card_group":"governance","is_public":0,"access_level":"auth","log_name":"policies","gen_nginx":1,"user_admins":"[]"}}