Conversation Analysis

Pattern analysis extracts signals from chat transcripts by identifying behaviors like frustration, success, and workflow transitions. The Skill processes multi-turn conversation logs and returns structured JSON with confidence scores, timestamps, and contextual quotes for each detected pattern.

Signal detection classifies user behaviors into taxonomy categories: success indicators, frustration points, workflow transitions, and tool-chain usage. The analysis applies across sessions to synthesize behavioral patterns and user preferences from the full conversation history.

Yes. After analyzing conversation history, the Skill generates syntheses and actionable recommendations in JSON format, including phase-driven workflow results, behavioral classifications with confidence levels, and contextual evidence from the original transcripts.

Conversation analysis accepts multi-turn chat transcripts and conversation logs as input. The Skill processes raw dialogue history to identify patterns, extracting structured signals without requiring pre-formatted data or external dependencies.

Conversation-specific analysis focuses on behavioral signals, workflow patterns, and user preferences across dialogue turns rather than static text content. It applies taxonomy-driven classification and detects temporal transitions between conversation phases for product insights.