loom-cli

Unofficial read-only CLI for Loom. Sign in once in Chrome. Then list your videos and pull official transcripts.

Built by True Frontier. Not affiliated with Loom or Atlassian.

The session stays on your machine, in a local Chrome profile. This tool does not print cookies, and it does not ship a cloud login.

Requirements

• Python 3.10+
• Google Chrome

Install

pipx install git+https://github.com/truefrontier/loom-cli.git

Or from a checkout:

pipx install .

Playwright uses your installed Chrome (channel="chrome"). You do not need playwright install chromium.

Commands

loom login
loom list
loom list --limit 20
loom list --id-only
loom list --compact | jq
loom list --select id,title --csv
loom info <id>
loom info <id> --format json --compact
loom transcript <id>
loom transcript <id> --format json
loom transcript <id> -o out.txt

loom login opens Chrome. Sign in, wait until your library loads, and the window closes. The profile is saved at ~/.loom-cli/chrome-profile/.

If a command says the session expired, run loom login again.

How it works

After login, the CLI calls the same GraphQL operations the Loom library uses (GetLoomsForLibrary, GetVideoCardDetails) from a signed-in Chrome profile. Transcripts are the official phrase file the share page already requests.

Read-only. It cannot upload, edit, delete, or share videos.

Agent output

Inspired by CLI Printing Press agent UX:

• Piped stdout auto-uses JSON when --format is table (no --json needed)
• --compact for high-gravity fields only
• --select id,title to project JSON/CSV fields
• --csv for spreadsheet-friendly rows
• --quiet / -q to suppress status lines
• Exit codes: 0 ok, 2 usage, 3 not found, 4 auth, 5 api/runtime

Limits

• Your library only. No public search.
• Transcripts exist only when Loom has generated one.
• A first page of the library is 12 videos. loom list follows the cursor until it hits your --limit, or until the library ends.