playbooks

home / skills / seefreed / skills / transcript-reflow

transcript-reflow skill

/transcript-reflow

This skill rewrites transcripts into clear, grammatically correct English with proper flow and preserved meaning, generating a new cleaned file.

npx playbooks add skill seefreed/skills --skill transcript-reflow

Review the files below or copy the command above to add this skill to your agents.

Files (1)
SKILL.md
1.3 KB
---
name: transcript-reflow
description: Clean and reorganize transcript .txt files for clearer English grammar, punctuation, and paragraph flow. Use when asked to fix line breaks, merge fragments into sentences, normalize punctuation, or produce a reflowed transcript saved as a new file.
---

# Transcript Reflow

## Overview

Rewrite transcript text into clear, grammatically correct English with sensible sentences and paragraphs while preserving meaning and stage directions.

## Workflow

### 1. Read the full transcript

Scan the entire file to understand topic flow, speaker intent, and where paragraphs should naturally break.

### 2. Reflow and rewrite

- Merge broken line fragments into full sentences.
- Normalize punctuation, capitalization, and spacing.
- Break long run-ons into readable sentences.
- Keep stage directions or cues in parentheses on their own lines.
- Keep the voice conversational; do not summarize or remove content.

### 3. Output rules

- Save the cleaned version to a new file with a clear name (e.g., `*_new.txt`).
- Keep ASCII-only text unless the source already contains non-ASCII.
- Preserve important proper nouns, model names, and product terms.

### 4. Quick self-check

- Read through to ensure paragraph transitions are smooth.
- Confirm no content was dropped or added.

Overview

This skill cleans and reorganizes plain-text transcript files into clear, grammatically correct English while preserving the original meaning and stage directions. It merges broken lines, normalizes punctuation and capitalization, and produces a reflowed version saved as a new file. The output is conversational and readable without summarizing or removing content.

How this skill works

The skill scans the entire transcript to understand speaker flow, intent, and natural paragraph breaks. It merges fragments into complete sentences, corrects punctuation and spacing, and breaks long run-ons into readable sentences while keeping stage directions or cues on their own lines. The cleaned transcript is written to a new ASCII-friendly file with a clear suffix (for example, *_new.txt). A final pass confirms no content was dropped or unintentionally altered.

When to use it

  • Fix transcripts with awkward line breaks from automated speech-to-text output.
  • Merge fragmented sentences from multi-line exports into fluid paragraphs.
  • Normalize punctuation and capitalization across a raw .txt transcript.
  • Prepare readable transcripts for publication, subtitles, or documentation.
  • Preserve stage directions while improving dialogue flow for editing or review.

Best practices

  • Provide the full transcript file so the skill can assess context for paragraphing.
  • Note any proper nouns, product names, or technical terms that must not be changed.
  • Specify if non-ASCII characters should be preserved or converted to ASCII.
  • Review the new file to confirm voice and intent match the original speakers.
  • Use the new file suffix convention (e.g., *_new.txt) to avoid overwriting the original.

Example use cases

  • Clean a meeting export that splits sentences at each speaker turn into a readable meeting summary.
  • Reflow interview transcripts for blog publishing while keeping interviewer cues intact.
  • Convert noisy speech-to-text output into punctuated dialogue for editing.
  • Prepare lecture transcripts for closed captions with clear paragraph breaks.

FAQ

Will the skill change the meaning of the transcript?

No. Edits focus on grammar, punctuation, and flow while preserving all original content and intent.

Can it handle non-English text or special characters?

Primary output is ASCII-only by default; retainment of non-ASCII is supported only if the source already contains them and you request it.

How is the cleaned file named?

The cleaned transcript is saved as a new file using a clear suffix like *_new.txt to avoid overwriting the original.