Manage Flags from Your Terminal: RocketFlag 2.11

RocketFlag 2.11 is live today, alongside the new RocketFlag CLI (@rocketflag/cli) built on top of the Management API we shipped in 2.9.

The short version of what shipped:

• The RocketFlag CLI (@rocketflag/cli), for creating, inspecting, toggling, and evaluating feature flags straight from your terminal or CI pipeline.
• API token improvements in the console and Management API so setting up tokens for the CLI and scripts is clearer and less error-prone.
• Stronger token and personal data cleanup when projects, organisations, or user accounts are deleted.

For now, Management API commands in the CLI (login, whoami, init, flags list, flags get, flags create, and flags toggle) authenticate with a project API token, so they are available on Teams and Enterprise (including the 14-day Teams trial). User-based login is on the roadmap, which will bring the full CLI to every RocketFlag user on every plan. In the meantime, evaluating flags with rocketflag flags eval calls the public Evaluation API and already works on every plan, including Solo.

The RocketFlag CLI (@rocketflag/cli)

Until today, adding a new feature flag meant leaving your editor, opening the console in a browser tab, creating the flag, copying its ID, and coming back to wire it into your code. Now you can do the whole loop from your terminal in a couple of commands.

Install the CLI globally with Node 22 or newer (or run commands directly via npx @rocketflag/cli):

npm install -g @rocketflag/cli

Once you have a project API token (rf_...) from the Tokens tab of your project, five commands take you from login to a live evaluation:

# 1. Authenticate and save your token to ~/.config/rocketflag/config.json
rocketflag login

# 2. Bind your repository to your project and default environment (.rocketflag.json)
rocketflag init

# 3. Create a flag enabled at 100% rollout in your default environment
rocketflag flags create checkout_v2 --enabled

# 4. Toggle the flag on or off (or pass --on / --off for idempotent CI scripts)
rocketflag flags toggle <flag_id>

# 5. Evaluate the flag via the public Evaluation API
rocketflag flags eval <flag_id>

A few details on how the CLI works:

• Credentials and repo config stay separate. rocketflag login validates your token against GET /api/v1/token before writing anything to disk and stores it in ~/.config/rocketflag/config.json with 0600 file permissions. rocketflag init writes .rocketflag.json in your repository with only { "projectId", "defaultEnv" }, never your secret, so you can commit it safely for the rest of your team.
• Scoped tokens for local development. If you create a local development token restricted to dev and test in the console, rocketflag whoami shows both the project’s full environment list and the environments your token is allowed to touch. Trying to toggle prod with that token fails immediately with a clear error message.
• Safe defaults on multi-environment projects. New flags default to a 100% rollout so flipping a toggle turns the flag on for all traffic unless you pass --percentage. When you pass initial state flags like --enabled, --cohort, or --audience on a multi-environment project, they apply only to your target environment (--env or defaultEnv), leaving your other environments safely switched off.
• Built for scripts and CI. Every command supports --json on stdout while keeping warnings and errors on stderr, and uses deterministic exit codes (0 for success, 1 for runtime/API errors, 2 for invalid usage or missing --env, and 3 for auth or permission failures).
• Test targeting rules from your shell. rocketflag flags eval <id> lets you pass --env, --cohort, --key (for sticky rollouts), and --attr key=value (for Audiences) to test how a flag evaluates without spinning up your app.

The new Getting Started with the CLI guide walks through each step on an existing project, and the CLI reference has the full option and exit-code tables.

API token improvements

Two changes to API tokens shipped in 2.11 to make setting up the CLI and CI scripts smoother:

• Explicit Read or Write selection. Previously, the Create Token modal in the console pre-selected Read permission by default, which made it too easy to create a read-only token by accident when setting up the CLI or a deployment script. You now have to pick Read or Write explicitly before creating a token.
• Project environments in token verification. Calling GET /api/v1/token now returns project.environments (the ordered list of environments defined on a multi-environment project, or [] on a single-environment project) alongside the token’s own allowed environments. That lets the CLI and your own scripts tell single-environment and multi-environment projects apart automatically and warn you if a token still references an environment label that was renamed or deleted.

The smaller things

Version label in the console header. Following the Carbon, Graphite & Brushed Silver redesign in 2.10, the current v{major.minor} release version (v2.11) is once again shown next to the RocketFlag wordmark in the top command bar.

Immediate token revocation on project or organisation deletion. Deleting a project or an entire organisation now immediately revokes all of its API tokens and clears their denial logs. Any script still calling the API with a token from a deleted project or organisation now receives a 401 Token revoked response straight away.

Cleaner personal data deletion. When someone deletes their RocketFlag user account, their personal feedback entries, dashboard snapshots, and pending invitations are deleted, while their past entries in organisation audit logs and any API tokens they created for a team are anonymised. That removes their personal details while keeping team API tokens and organisation audit trails intact.

Give it a spin

If you’re on Teams or Enterprise (or on a 14-day Teams trial), grab a token from Project > Tokens in the console and run npx @rocketflag/cli login. If you’re new to RocketFlag, the Solo tier is free without a credit card, and Teams is a flat $79/month for your whole organisation. Other tools’ bills go up when you hire, or when your users show up; RocketFlag’s doesn’t.

If you run into anything rough with the CLI, or have a command you’d like to see added next, contact us and let me know.