video-enhancer

name: video-enhancer description: >- Video enhancement pipeline. Use this skill when the user wants to sharpen, colour grade, warm, or normalise the audio of videos — including presets for natural, cinematic, or vivid looks. metadata: { "openclaw": { "emoji": "🎨", "requires": { "bins": ["ffmpeg", "uv"] }, },

}

video-enhancer — Video Enhancement Pipeline

Sharpens, colour-grades, and normalises audio for a batch of videos. No captions — for captioning use the video-captioner skill.

Pipeline

input video → unsharp + eq + colorbalance (ffmpeg) → -14 LUFS audio normalisation → output

Presets

If the user has not specified a preset, ask them which look they want before running. Describe the options briefly: natural for a clean, understated result; cinematic for a warm, punchy Instagram-ready grade; vivid for maximum colour intensity.

How to run

Install dependencies (first run only):

cd "$SKILL_DIR" && uv sync

Then process videos:

cd "$SKILL_DIR" && uv run python scripts/enhance.py \
  --input <path/to/input> \
  --output <path/to/output> \
  --preset natural|cinematic|vivid

Common invocations

# Cinematic grade
uv run python scripts/enhance.py --input ./input --output ./output --preset cinematic

# Natural grade
uv run python scripts/enhance.py --input ./input --output ./output --preset natural

# Vivid grade
uv run python scripts/enhance.py --input ./input --output ./output --preset vivid

After running

Report back to the user:

• How many videos were processed successfully and how many failed.
• The full path to the output directory.
• If any videos failed, name them explicitly.

Edge cases

• If ffmpeg is not installed: tell the user to install it (sudo pacman -S ffmpeg on Arch/CachyOS, brew install ffmpeg on macOS, sudo apt install ffmpeg on Ubuntu)
• If uv is not installed: direct to https://docs.astral.sh/uv/getting-started/installation/
• If the input directory is empty: report clearly rather than silently exiting
• If an unknown preset is given: list the valid options (natural, cinematic, vivid)
• Audio normalisation may print a warning about dynamic mode for short clips — this is expected and not an error