Render CLI for power users

Prerequisites

• A Render account with at least one service or database you can deploy against
• Comfort with the terminal, environment variables, and a Unix-ish shell (bash or zsh)
• Basic familiarity with `jq` is helpful but not required

Steps

1. 01 Why the CLI, and what you'll build Map the Render Dashboard-to-script maturity ladder, set expectations, and preview the GitHub Actions capstone you'll ship at the end. 4 min →
2. 02 Install and pin the CLI Install the Render CLI on any platform, verify the version, and pin it for reproducible CI runs. 5 min →
3. 03 Authenticate and switch workspaces Master the two auth modes (login token vs API key), the precedence rules between them, and the workspace-switching patterns that keep multi-account work sane. 8 min →
4. 04 Output formats and non-interactive mode Internalize the TTY auto-detect rules, the `-o` flag, the `RENDER_OUTPUT` env var, and `--confirm` - the four levers every script depends on. 8 min →
5. 05 List, filter, and pipe with jq Turn `render services` into a Swiss-army knife. Get service IDs by name, filter by type and status, and build the lookup helpers every later step depends on. 9 min →
6. 06 Trigger deploys you can trust Master `render deploys create` - from a plain trigger to deploying a specific commit, a specific image, waiting on completion, and reading the exit code in a script. 10 min →
7. 07 Tail, filter, and triage logs Stream logs from your terminal, scope them to a service or deploy, pipe them through `grep` and `jq`, and capture them on failure. 9 min →
8. 08 SSH, ephemeral shells, and psql Get a shell on a running instance, spin up a clean throwaway one for risky pokes, and run one-shot SQL queries against your Postgres database from any script. 11 min →
9. 09 Blueprint validation and AI tool integration Lint `render.yaml` from your terminal and CI, and install Render skills so your AI coding tools speak Render fluently. 8 min →
10. 10 Capstone - deploy on push, fail loudly, dump logs Wire every pattern from the course into a real GitHub Actions workflow that deploys on push, exits non-zero on failure, and dumps the last 200 log lines so you can debug from the PR. 12 min →