summarize-interview

This skill is designed for product managers, user researchers, and startup founders who need to extract actionable insights from raw customer interview transcripts. It acts as a bridge between unstructured qualitative data and structured product discovery documentation. By parsing interview recordings, transcripts, or notes, the agent identifies core customer pain points, existing solutions, and specific Jobs to Be Done (JTBD) using established product discovery frameworks.

• Automatically processes interview content to extract key themes such as customer background, currently used solutions, and satisfaction levels regarding those solutions.

• Employs a structured template format to ensure consistency across multiple customer sessions, making it easier to compare findings across different user segments.

• Identifies 'satisfaction signals' that help product teams distinguish between high-priority problems and minor inconveniences.

• Extracts and formats clear, time-bound action items with assigned owners, facilitating immediate follow-up and implementation.

• Translates complex interview nuances into plain, clear language that is accessible to cross-functional stakeholders including designers, engineers, and executives.

• Input: Users can provide data via attached files (PDF, text, transcript) or direct text input. The agent works best when provided with full transcriptions from discovery interviews or research sessions.

• Output: A markdown-formatted summary report that is ready to be saved in a workspace or knowledge base for team review.

• Usage Tips: This skill is a primary building block for the 'Product Discovery' workflow. To achieve the best results, ensure the transcript includes clear dialogue between the interviewer and the participant. The agent will replace numeric scores with qualitative descriptions to maintain a human-centric focus on the customer experience.

• Constraints: The effectiveness of the summary is directly tied to the clarity of the source transcript. If information is missing from the recording, the template will explicitly note this with a hyphen, ensuring the data remains accurate rather than hallucinated.