transcript-polisher_skill

Overview

This skill transforms raw podcast and interview transcripts into polished, publication-ready documents while preserving the speaker's authentic voice and key insights. It aggressively removes filler, fixes grammar, and adds clear structure and timestamps so spoken content reads like a clean article or show notes. The result is shorter, more readable text that stays faithful to meaning.

How this skill works

The skill applies a multi-step workflow: add a standardized header and a timestamped outline, replace generic speaker labels with bolded names, and run an aggressive editing pass that eliminates filler words and false starts. It preserves powerful phrasing, emotional moments, and technical terms while trimming redundancy to a target 25–35% length reduction. Final quality checks ensure speaker consistency, accurate timestamps, and publication-ready markdown output.

When to use it

• Cleaning automated transcripts from tools like Otter, Rev, Descript, or YouTube captions
• Turning interview or panel audio into readable show notes or long-form articles
• Preparing spoken content for publication without altering meaning or voice
• Converting messy conversational recordings into structured chapters for repurposing

Best practices

• Keep original impactful quotes and technical terms exactly as spoken
• Bold speaker names and use consistent first-name or full-name formatting
• Create up to 10 compelling timestamped chapters for 45–60 minute episodes
• Remove all filler words and consolidate repeated or false-start passages
• Aim for a 25–35% length reduction while running a final fidelity checklist

Example use cases

• Publishable podcast transcripts with chaptered markdown and show notes
• Edited interview transcripts for newsletters, blog posts, or long-form articles
• Internal highlights and timestamped summaries for team distribution
• Repurposing podcast content into social posts or episode summaries