Coherent Transcript Editor

Faithful, coherent editing of a raw spoken-transcript (auto-transcribed, timestamps/fillers/errors) into a clean readable document - WITHOUT summarizing, paraphrasing, or unifying speakers into one voice. "Every word verbatim" editing for DIALOGUE / PANEL / INTERVIEW / DEBATE / Q&A: keep every word, idea, story and aside of EVERY speaker in their own first person; only fix punctuation, paragraphing and transcription garble, remove timestamps/stutters, add bold speaker labels + topic headings, restore quoted verses to exact text + source, then build an RTL Hebrew DOCX. TRIGGER on Hebrew: 'עריכה קוהרנטית', 'ערוך תמלול דו-שיח', 'ערוך תמלול ראיון', 'ערוך תמלול פאנל', 'כל מילה ככתבה וכלשונה'. English: 'coherent edit of transcript', 'faithful transcript editing', 'edit this dialogue/interview transcript'. USE for a multi-speaker (or faithful single-speaker) transcript that must stay VERBATIM. NOT for a talk rewritten in first person, NOT for summaries.

hiyabh Updated

File contents

hiyabh/claude-skills/tree/main/src/coherent-transcript-editor commit 00cd536feb

Frequently asked questions

npx skillmds@latest add hiyabh/coherent-transcript-editor