Render CLI

Render CLI

Use the render binary to manage Render services, deploys, logs, and sessions from the terminal. Treat it like a production-control CLI: read first, parse JSON, and only create or mutate resources when the operator explicitly asked for that action.

Agent defaults

Always add these flags for agent-safe commands:

• --output json for parseable output.
• --confirm to skip confirmation prompts.

The CLI defaults to interactive mode and can hang an agent if a command prompts. If a command is intended to run unattended, include both flags unless the CLI help says the mode is interactive-only.

Auth

Check the active login before assuming access:

render whoami --output json --confirm

If the user is not logged in, ask them to complete:

render login

render login uses the dashboard/browser login flow, so do not run it in an unattended automation loop.

Services

List services in the active workspace:

render services --output json --confirm

Use the returned service id as <service-id> in deploy and log commands. When a push to Git did not deploy, inspect the service JSON for autoDeploy; it may be disabled for that service.

Deploys

Trigger a deploy only when the operator explicitly asks for a real deploy:

render deploys create <service-id> --output json --confirm

Watch deploy state by polling:

render deploys list <service-id> --output json --confirm

Common deploy statuses include queued, build_in_progress, update_in_progress, live, and failed. Treat live and failed as terminal states for ordinary deploy monitoring.

Logs

Query recent logs for a service:

render logs --resources <service-id> --limit 100 --output json --confirm

For a live tail, use:

render logs --resources <service-id> --tail --confirm

The CLI help says --tail streams only in interactive mode, so prefer bounded JSON log queries for automated agents and use live tail only when a long-running interactive stream is acceptable.

Gotchas

• The CLI is interactive by default; --confirm prevents confirmation prompts from blocking the agent.
• --output json makes service, deploy, and log output machine-readable.
• In non-interactive log mode, --resources <service-id> is required.
• Git pushes do not guarantee deploys. Check autoDeploy in the service JSON before assuming push-to-git should deploy.
• Do not use placeholder IDs in real commands. Resolve a service ID from render services --output json --confirm first.