diagnosis-agent

To diagnose marketing campaign performance, this Skill analyzes calculated_metrics.json, buyer_persona.json, and experiments_log.json to produce a structured diagnosis and optimization plan segmented per persona_id and channel.

Creative fatigue detection identifies underperforming ad assets by analyzing knowledge patterns and metrics, triggering proposed closures and new experiment hypotheses within the optimization_actions.json output for weekly tracking.

Generating an optimization plan requires inputs like brand_strategy.json and knowledge_base_marketing.json, which the Skill processes to output 2-3 weekly optimization actions, proposed experiments, and structured escalation risks.

Yes, buyer persona segments are applied per persona_id across channels, using calculated_metrics and strategy context to prescribe targeted optimization actions and assign suggested_owners with supporting evidence.

You need to provide JSON data formats including calculated_metrics, knowledge_base_marketing, buyer_persona, brand_strategy, and experiments_log to generate diagnosis.json and optimization_actions.json outputs.

Marketing experiments should be closed when sufficient evidence is gathered, which the Skill determines by analyzing experiment logs and metrics to populate proposed_closures with decision_level and problem_domain attributes.