Lizard CLI

Deploy apps, databases and sandboxes to Lizard (lizard.build) from the command line.

Point the CLI at your code and it detects the stack — Go, Node, Python, Rust, Ruby, PHP, Java or a static site — builds it and ships it live. Managed Postgres, Managed Redis and Managed Object Storage attach in one command. Billing only for active resources, so idle apps cost a fraction.

Every command is agent-readable: add --json for structured output, or run lizard --help --json to dump the full command schema.

Install

curl -fsSL https://lizard.build/install.sh | bash

On Windows (PowerShell):

irm https://lizard.build/api/cli/install.ps1 | iex

Or via npm:

npm i -g @lizard-build/cli

Start here (AI agents)

lizard skills get core

Skills ship inside the CLI, so the guide always matches the installed version. They cover the whole app lifecycle — deploy, link, addons, logs, scaling, secrets, domains — with copy-paste examples. Prefer this over guessing commands from flag docs alone.

Usage

lizard login
lizard init
lizard up

lizard login signs you in and offers Pro if the account has no plan. The install script runs it for you, and any command that needs an account signs you in on the spot. On a server or in CI, set LIZARD_TOKEN, or save a token with printf '%s' "$TOKEN" | lizard login --token-stdin.

lizard up uploads the code, builds it and returns a live URL on onlizard.com. Add managed data services in one command:

lizard add postgres redis s3

Common commands

lizard up         # upload and deploy code
lizard add        # add Managed Postgres, Managed Redis or Managed Object Storage
lizard ps         # list services
lizard status     # show the linked workspace, project and service
lizard logs       # stream runtime logs
lizard metrics    # resource metrics (CPU, memory, network, disk); --cost for spend
lizard billing    # plan, tier, trial, usage this month; start, change tier, cancel or resume Pro
lizard metrics --all # all services in the project, even with a linked service
lizard scale      # scale replicas / CPU / memory / storage
lizard pause      # stop a service (app or database) without deleting it; `lizard resume` starts it again
lizard project pause   # pause every service of a project (also: project resume, workspace pause/resume)
lizard workspace resale-pricing <ws> --all 3   # resellers: the workspace's scoped keys see 3x Lizard's prices (your bill is unchanged)
lizard workspace limits <ws> --sandboxes 5 --cpu 2 --memory 4GB --disk 20GB   # cap a workspace: sandboxes at once, CPU/RAM its services use, SSD its data holds
lizard sleep --on # serverless: sleep when idle, wake in ~1 s on the next request (--off: always on)
lizard secrets    # manage secrets
lizard domain     # manage domains
lizard redeploy   # rebuild and redeploy from the current source
lizard run        # run a command locally with project and service secrets injected
lizard ssh        # disabled for services (403); use `lizard sandbox ssh` for sandboxes
lizard sandbox    # create and manage Sandboxes
lizard sandbox create --size large # 8 vCPU / 16 GB; small, medium (default), large
lizard sandbox desktop <id> --open # graphical desktop in the browser (sandbox create -t desktop)
lizard sandbox vnc <id>            # the same desktop in a native VNC app (localhost:5900)
lizard sandbox ssh <id>            # interactive shell in any running sandbox (no keys to set up)
lizard agents                      # coding agents: connected credentials and which agents can run
lizard agents login chatgpt        # ChatGPT subscription (device code); `agents add anthropic|openai|openrouter|claude …` for keys/tokens
lizard sandbox prompt <id> "…"     # run Codex / Claude Code / pi / OpenCode / Prime in the sandbox, streaming (--agent)
lizard volume     # Persistent Volumes for Sandboxes (create / list / resize / rm)
lizard skills     # agent guides, version-matched to the CLI

run executes on your machine with the service's stored secrets injected; pass an addon (-s postgres) to get its real connection values. Shell access into services is off, so lizard ssh answers 403. To check a service's variables, read lizard logs --build -s my-app (it warns about every empty reference) and lizard secrets list -s my-app --show.

lizard run -s postgres -- npm run migrate

Run lizard --help for the full list.

Billing

lizard billing          # plan, trial, usage this month, next charge
lizard billing start    # start the Pro trial in Stripe Checkout
lizard billing tier 5x  # move a paid subscription to pro, 5x or 20x
lizard billing promo X  # redeem a promo code for a longer trial

Pro costs $19/month, taxes included, with $19 of usage each month. Usage above that is pay as you go. A new account starts with no plan; the 7-day trial needs a card, includes $5 of usage and charges nothing until it ends. When the account has no plan, commands that create something fail with the reason and the link to open next.

For agents and scripts

Add --json to any command for structured output. The CLI also switches to JSON on its own when stdout is not a TTY. Errors come back as {"error": {"code", "status", "message", "body"}} and exit codes are stable:

0  success          3  not found (404)
1  generic error    4  timeout (408/504)
2  auth (401/403)   5  cancelled by user

lizard --help --json dumps the machine-readable schema for any command — arguments, options, defaults, subcommands and exit codes.

Links

• Docs — https://lizard.build/docs
• Source — https://github.com/lizard-build/lizard-cli

License

MIT