Can it handle technical jargon or specialized terminology?

Yes. The editing guidelines explicitly state not to edit specialized terminology or cultural references, ensuring industry-specific context is preserved.

Does this skill change the original meaning of the transcript?

No. The skill follows a strict 100% fidelity rule, ensuring that while filler words and redundancies are removed, the core ideas, specific insights, and powerful statements remain untouched.

What format is the final output provided in?

The polished transcript is delivered in clean Markdown, featuring a timestamped outline, bolded speaker names, and structured headers for maximum readability.

How much does the skill reduce the length of a transcript?

On average, the skill achieves a 25-35% reduction in length by eliminating verbal hedging, false starts, and redundant explanations while keeping the strongest examples.

Is this skill suitable for editing pre-written articles?

No. This skill is specifically optimized for raw spoken content that contains the natural dysfluencies of conversation; it is not intended for content that is already polished or scripted.

Transcript Polisher

Name: Transcript Polisher
Author: cdeistopened

bycdeistopened

•

コンテンツ管理

Transforms raw, messy transcripts into polished, professional documents while preserving the speaker's original voice and meaning.

Transcript Polisher is a specialized capability for Claude Code designed to convert automated transcripts from services like Rev, Otter, or Descript into high-quality, readable content. It aggressively removes filler words, fixes grammatical errors, and structures text with timestamped chapters and clear speaker identification. By targeting a 25-35% reduction in length through the elimination of redundancies and false starts, it ensures the final document is concise and ready for publication as show notes or articles without sacrificing the unique personality or intended meaning of the speakers.

主な機能

01Redundancy consolidation while maintaining 100% fidelity to meaning

02Automatic filler word removal (um, uh, you know, like)

03Grammar and sentence fragment repair for professional readability

04Accurate speaker identification and consistent bold formatting

051 GitHub stars

06Timestamped chapter generation and document structuring

ユースケース

01Refactoring auto-generated video captions into readable articles

02Converting raw podcast transcripts into polished show notes or blog posts

03Cleaning up interview recordings for professional research or documentation

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add cdeistopened/skill-stack transcript-polisher

For use in Claude.ai and ChatGPT

Download Skill