Prompt	Entity	Model	Response Preview	TokensⓘHow many words the AI used in its response. More tokens = more detailed answer.	LatencyⓘHow long the AI took to respond. Faster models answer in under 5 seconds.	Cost	Status

Prompt	Target	Model	SentimentⓘOverall tone of what AI said — positive, neutral, or negative.	SN?ⓘWas any Shepherd Neame entity mentioned in the response?	Rec?ⓘWas the entity actively recommended, not just listed?	Brand?ⓘDid AI name Shepherd Neame alongside the pub or beer?	Mentions	Entities Found

Run	Status	Models	OK	Fail	Poll CostⓘEstimated cost of sending prompts to the AI models.	Parse CostⓘEstimated cost of analysing the responses with Claude.	TotalⓘCombined cost of polling and parsing for this run.	Duration	Date