@loomta/cli

Loomta CLI — the cheapest way to grow your app on TikTok. Schedule and publish TikTok posts (video + photo slideshows), generate captions, and manage media from the command line or an AI agent.


Keywords
loomta, tiktok, social-media, scheduling, automation, ai-agent, cli
License
MIT
Install
npm install @loomta/cli@0.1.1

Documentation

Loomta CLI

The cheapest way to grow your app on TikTok — from your terminal or an AI agent.

loomta is a thin, JSON-in/JSON-out command-line wrapper over the Loomta public API. It lets you (or an AI agent) schedule and publish TikTok posts, upload media, and generate captions without touching the web app.

Install

npm install -g @loomta/cli
# or
pnpm install -g @loomta/cli

Requires Node.js ≥ 20.

Authenticate

Create an API key in your Loomta workspace settings (API Keys → Create key), then:

loomta auth:login --api-key loomta_sk_xxx
# or
export LOOMTA_API_KEY=loomta_sk_xxx

loomta auth:status

Credentials are stored at ~/.loomta/credentials.json (mode 600). The LOOMTA_API_KEY environment variable takes priority. Point at a self-hosted backend with LOOMTA_API_URL.

Quick start

# Find your TikTok platform id
loomta platforms:list

# Upload media (required before posting — returns a READY object with an id)
MEDIA_ID=$(loomta upload launch.mp4 | jq -r '.id')

# Publish now
loomta posts:create -c "We just shipped 🚀 #buildinpublic" \
  -m "$MEDIA_ID" -i "<platform-id>" --mode now

# Or schedule
loomta posts:create -c "Coming soon" -m "$MEDIA_ID" -i "<platform-id>" \
  -s "2026-07-01T12:00:00Z"

Hands-off photo slideshow (AI writes captions, generates images, and creates the post):

# --wait polls the async job until it finishes and prints the final result
loomta slides:generate -i "<platform-id>" --mode draft \
  --topic "3 reasons indie devs love our app" --wait

Commands

Command Description
auth:login --api-key <key> Validate and store an API key
auth:status Check the current key
auth:logout Remove stored credentials
platforms:list List connected accounts and their IDs
platforms:schema <id> Posting limits + settings for a platform
upload <file> Upload an image/video, returns a READY media object
media:list List uploaded media (--folder-id, --media-type, --source)
media:folders:list / media:folders:create <name> Manage folders
posts:create Create a post (-c content, -i platform ids, -m media ids, --mode, -s, --comments)
posts:list List posts in a date range
posts:get <id> Fetch one post group
posts:reschedule <id> -s <iso> Change scheduled time
posts:publish <id> Publish or schedule a DRAFT (--mode now|schedule, -s)
posts:delete <id> Delete a post
slides:caption Generate AI captions for a slideshow (uses credits)
slides:composite Bake captions onto images (free)
slides:generate Hands-off slideshow: captions + AI images + post, async (--wait to poll)
slides:generate:status <jobId> Check a generation job's status

Both posts:create and posts create syntaxes work. Run loomta <command> --help for details.

Use as an AI agent skill

This repo ships a SKILL.md and a .claude-plugin manifest so agents can install and drive the CLI directly. See SKILL.md for the full agent playbook.

License

MIT