loom-analyze — requirements & troubleshooting

Detail split out of SKILL.md per progressive disclosure. Read this only when setup fails or the user hits a limitation.

Install (per-machine, one-time)

Copy this entire folder to ${CLAUDE_PLUGIN_ROOT:-$HOME/.claude}/skills/loom-analyze/ on the target machine.
Run setup: bash ${CLAUDE_PLUGIN_ROOT:-$HOME/.claude}/skills/loom-analyze/setup.sh — installs Homebrew packages yt-dlp + ffmpeg and pip package openai-whisper. Idempotent.
Restart Claude Code. The skill is auto-discovered from the SKILL.md frontmatter.

Requirements

macOS (or Linux with Homebrew). Windows + WSL works the same.
Python 3 (for whisper).
Disk for the whisper model, downloaded on first run: base ~140 MB, small ~460 MB, medium ~1.5 GB, large ~3 GB.

Troubleshooting

“Missing dep” error: re-run setup.sh.
whisper not found after install: ensure ~/Library/Python/3.x/bin (or wherever pip user-installs binaries) is on PATH.
Loom URL not recognised: must be loom.com/share/{id}. Embedded/private links won’t work without auth.

What it does NOT do

No streaming / real-time. Whole-video transcription only.
No speaker diarization (single transcript stream).
No upload to any third-party service. All processing is local.