What problem does it solve? Raw reply rate hides whether a cold email campaign is actually working — a 5% reply rate full of unsubscribes and hostile replies is worse than a 1% reply rate from interested buyers. This Skill classifies every reply from a Smartlead campaign and computes the positive reply rate, the metric that shows whether a campaign generates real demand. ## Core Features & Use Cases - Reply Fetching: Pulls all leads with replies and their first inbound message from the Smartlead API, with pagination, retry logic, and optional date filtering. - Claude-Based Classification: Labels each reply as positive_interested, positive_soft, positive_referral, neutral, negative, unsubscribe, OOO, bounce, or other using a tunable prompt. - Rate Aggregation & Benchmarks: Computes positive reply rate, hostile rate, and unsubscribe rate against B2B cold email benchmarks, excluding OOO and bounces from denominators. - Use Case: After a campaign has run for two weeks, score its replies to decide whether to scale it, kill it, or run a deliverability audit — and surface the positive replies that need an immediate human response. ## Quick Start Score the replies for Smartlead campaign 12345 and tell me the positive reply rate plus which leads need a response right now.
