audio-review
/audio-review
Judge audio like a senior audio designer (loudness, dynamics, tonal balance, stereo, noise, artifacts, delivery specs) and give a scored review with fixes. Use when the user shares audio or a video's sound (podcast, voiceover, music, mix, wav, mp3) and asks how it sounds, whether it is ready, or for an audio review.
- Author
- Shash-Builds ↗
- Repository
- Shash.builds ↗
- Licence
- MIT
- Last repository change
- 07.11.2026
Assessment and security
- Files in this package
- 7
- Scripts
- 2
- HTML comments
- 0
- URLs in SKILL.md
- 0
No signals detected. Repository changed on 07.11.2026, without references to retired models or APIs.
MIT, open: use and modify while retaining author attribution.
- 2 scripts in this package. Review them before installing, as Claude may execute them.
Scripts: analyze.py, review.py
What it does and when to use it
Judge audio like a senior audio designer (loudness, dynamics, tonal balance, stereo, noise, artifacts, delivery specs) and give a scored review with fixes. Use when the user shares audio or a video's sound (podcast, voiceover, music, mix, wav, mp3) and asks how it sounds, whether it is ready, or for an audio review.
Original author description
Judge audio like a senior audio designer (loudness, dynamics, tonal balance, stereo, noise, artifacts, delivery specs) and give a scored review with fixes. Use when the user shares audio or a video's sound (podcast, voiceover, music, mix, wav, mp3) and asks how it sounds, whether it is ready, or for an audio review.
How it works
- Steps
- Other ways to run it
SKILL.md preview
---
name: audio-review
description: Judge audio like a senior audio designer (loudness, dynamics, tonal balance, stereo, noise, artifacts, delivery specs) and give a scored review with fixes. Use when the user shares audio or a video's sound (podcast, voiceover, music, mix, wav, mp3) and asks how it sounds, whether it is ready, or for an audio review.
---
# Audio review
You can't hear audio directly. Instead, measure the file the way an engineer
reads meters and analysers while listening, then judge it as an audio
designer.
All the files this skill uses sit next to this SKILL.md. In Claude Code that
folder is `${CLAUDE_SKILL_DIR}`. Anywhere else, use the folder this file was
loaded from. Below, `SKILL_DIR` means that folder.
## Steps
1. **Find the file(s).** Get the audio path from the user, plus a reference
track if they gave one. Ask what it's for (streaming music, podcast,
broadcast, film, game, ad, social, UI sound) only if you can't reasonably
infer it.
2. **Analyse.** Write the output to a scratch or temp directory:
```bash
python3 "SKILL_DIR/analyze.py" "<file>" --out "<out_dir>"
```
If you have a reference track, run the same command on it with its own
`--out`. This needs `ffmpeg` and `numpy`. If they're missing, install them
(`pip install numpy`, plus ffmpeg from the system package manager). If you
can't run code at all, skip to the no-code fallback in
`SKILL_DIR/initiation_prompt.md`.
3. **Look and read.** Read `<out_dir>/metrics.json`. Then open each picture so
you actually see it: `spectrogram.png`, `spectrogram_linear.png` and
`waveform.png`. Check the analyser's `flags` against the pictures; some are
heuristics.
4. **Judge.** Read `SKILL_DIR/designer_rubric.md` and follow it exactly: the
same voice, reference points and report format (verdict and score, what
I'm hearing, scorecard, ranked issues with evidence, fix list with starting
values, what's working, limits of this review).
5. **Reply** with the review. If the user wants a file, also save it as
`<out_dir>/review.md`.
## Other ways to run it
- `SKILL_DIR/review.py` does the same thing through the Claude API, for
scripts and batch jobs.
- `SKILL_DIR/initiation_prompt.md` is a paste-in prompt for any Claude chat.
See `SKILL_DIR/README.md` for details.
Files in this package (7)
- README.md 3 KB
- SKILL.md 2 KB
- analyze.py 26 KB
- designer_rubric.md 5 KB
- initiation_prompt.md 5 KB
- requirements.txt 1 KB
- review.py 5 KB
Versions
The database has no other skills named audio-review.
Database snapshot: 13.11.2026 · package version 93a765b0484b (SHA-256 digest of the entire package)
Downloads in the last 30 days: 0 · ranking
Report a problem with this skill: harmful content, incorrect descriptions or an author’s removal request.
Similar skills
- 100sound-designer
The studio's Sound Designer. Owns the sound of a video for the studio-director, covering dialogue cleanup, music bed and ducking, sound effects synced to picture and motion, the final mix, loudness per platform, and stems, and reports back with a Sound Handoff. Picks and runs… - 100audio-mixing
Mix voiceover, background music and sound effects for video - ducking music under speech, fades, loudness normalization (LUFS), noise reduction and music selection. Use when audio sounds uneven, music overpowers speech, or the user asks to add or balance music. - 100audio-verification
Check exported audio with measurements instead of guessing, since the agent can't listen. Use after rendering a mix or stem from FL Studio, when the user reports noise, clicks, thumps, harshness or low volume, to confirm that chords sound the notes that were written, to prepare… - 100elevenlabs
Generate and transcribe audio through ElevenLabs' HTTP API — text-to-speech voiceover, multi-speaker dialogue, sound effects, background music, and transcription. Use when the user asks for a voiceover, narration, an AI voice, podcast or audiobook audio, a sound effect,… - 100elevenlabs
Generate and transcribe audio through ElevenLabs' HTTP API — text-to-speech voiceover, multi-speaker dialogue, sound effects, background music, and transcription. Use when the user asks for a voiceover, narration, an AI voice, podcast or audiobook audio, a sound effect,… - 88stable-audio
Generate and finish local music and sound effects with Stable Audio 3 through the stable_audio MCP server, including candidate comparison, audio inspection, loops, fades, loudness and WAV/OGG/FLAC export. Use for local music/SFX production; AI continuation, inpainting and singing are unavailable.