auto-editor-transcribe

An agent skill by WyattBlue, from WyattBlue/auto-editor. Tags: audio, cli, editing, subtitles, transcription, video.

What it does

Transcribe a media file or live microphone to text or subtitles with auto-editor's whisper subcommand, then edit by spoken content — cut to keep (or drop) sections whose subtitles match a word or regex. Use when someone wants a transcript, captions, live speech transcription, or to cut a video based on what is said rather than on loudness.

Install

With the skills CLI, which installs into Claude Code, Codex, Cursor and other agents:

npx skills add WyattBlue/auto-editor --skill auto-editor-transcribe

Or copy the skill folder into Claude Code's skills directory by hand (~/.claude/skills for every project, or .claude/skills inside one):

git clone --depth 1 https://github.com/WyattBlue/auto-editor
cp -r auto-editor/skills/auto-editor-transcribe ~/.claude/skills/auto-editor-transcribe

Safety box score

Grade B, low risk: 14 of 100 risk points, rated 2026-09-26. The grade rates what the skill and its scripts can reach on the machine of whoever installs it, across eight categories from shell execution to secrets access. How the score works.

This skill uses the local `auto-editor` CLI to transcribe media or live microphone audio and then cut media based on subtitle/word matches. The main risks are model downloads over the network, live microphone access, and writing/choosing transcript or edited output files in the local workspace.

Category levels, from 0 (none) to 3 (high)
Category	Level	Why
Code & shell execution	1 (low)	It explicitly runs local commands like `auto-editor whisper <file\|:mic> <model>` and `auto-editor subdump FILE`, but it does not ship a script or code-eval payload.
Network access	1 (low)	The skill points to downloadable models on HuggingFace and says Apple speech "may download Apple's speech model and therefore needs network access."
Filesystem reach	1 (low)	It writes transcript outputs with `-o out.srt` and can auto-load sibling `talk.srt`/`talk.ass`, so it reads/writes nearby files but stays local.
Secrets & credentials	0 (none)	No credential stores, API tokens, passwords, `.env` files, keychains, or browser sessions are mentioned.
Privilege & persistence	0 (none)	There is no sudo, service installation, global package install, or shell/OS setting change described.
Instruction hijack surface	0 (none)	The regex examples are just subtitle matching; nothing instructs the agent to ignore instructions, fetch external prompts, or hide behavior.
Irreversible actions	1 (low)	Its purpose is to cut timelines based on speech with `--edit subtitle`/`word`, which changes media content, but it does not publish, send, or deploy anything.
Transparency	0 (none)	The skill is plain markdown with readable commands and no binaries, base64 blobs, or hidden runtime code.

Before installing

Confirm whether you are okay with first-run model downloads from HuggingFace or Apple.
Run `auto-editor whisper` on a non-sensitive sample file first and check where `-o` writes the transcript.
If you plan to use `:mic`, make sure you want live microphone input captured and transcribed.
Check for existing sibling subtitle files like `talk.srt`/`talk.ass` so they do not get auto-selected unexpectedly.

Rated 2026-09-26 by an automated review of 1 file from the skill's repository.

Source

Repository: WyattBlue/auto-editor (all skills from this repository)
Path: skills/auto-editor-transcribe/SKILL.md
Branch: master
Updated: 2026-09-23

auto-editor — Record a microphone or edit video/audio with the auto-editor CLI — cut out dead space (silence, motionlessness, black frames), set the pace, make manual cuts.
auto-editor-export — Export auto-editor's edit or microphone recording to a video editor (Premiere, DaVinci Resolve, Final Cut Pro, ShotCut, Kdenlive), to individual clips.
auto-editor-effects — Apply actions and animations to sections of a video/audio with auto-editor — change speed, duck or fade volume, de-ess, zoom/Ken Burns, rotate/spin.
youtube-downloader — Download YouTube videos and HLS streams (m3u8) from platforms like Mux, Vimeo, etc. using yt-dlp and ffmpeg.
add-dictation — Use when the user runs /add-dictation or wants speech turned into text with Grok speech-to-text: a mic button that dictates into the composer, live captions.
asr-transcribe-to-text — Transcribe audio/video to speaker-labeled text — who-said-what by default, plain-text opt-out; MLX-local on Apple Silicon or remote; local files, media URLs.