Commands

Command reference for raptor, the kubectl-style Facets CLI: verbs and resource groups for projects, blueprints, releases, artifacts, and IaC modules.

Raptor is a kubectl-style CLI. Commands are organized into action verbs (get, create, apply, set, delete, describe, plan) acting on resource categories.

For the authoritative command reference, see Facets-cloud/raptor-releases. For inline, in-terminal help:

raptor --help                  # Top-level command list
raptor <command> --help        # Flags and examples for any subcommand
raptor blueprint-guide         # Embedded design guide for blueprints

Command groups

Category	Purpose	Examples
Authentication	Log in and inspect identity	`login`, `whoami`, `auth can-i`
Projects & types	Manage projects and project types	`get projects`, `create project`, `get project-types`, `import project-type`
Resources (blueprint)	Add and update resources	`apply resource`, `get resources`, `delete resource`
Overrides	Per-environment configuration	`apply override`, `get overrides`, `delete override`
Plan & deploy	Validate and roll out changes	`plan`, `create release`, `logs release`, `abort release`
Environments	Provision and tear down environments	`create environment`, `launch environment`, `destroy environment`
Artifacts & builds	Register container images and zip bundles, and manage what is registered	`create artifact`, `set artifact-uri`, `set artifact-zip`, `get artifacts`, `get builds`, `promote build`, `delete build`
Delivery configuration	Decide where a build lands: branch mapping and promotion ladder	`get ci-cd`, `set ci-cd`, `test ci-cd`, `attach artifact`, `detach artifact`, `get artifact-usages`
Registries	Container registry credentials	`get registries`, `get registry-credentials`
Modules (IaC)	Author and publish Terraform modules	`module init`, `module set-spec`, `create iac-module`, `publish iac-module`
Variables & secrets	Project configuration values	`get variables`, `create variable`, `set variable`
Discovery	Schema and expression lookup	`describe module`, `describe expressions`, `get resource-types`
Impact	Dependency analysis	`impact <kind>/<name>`, `impact ... --reverse`
Users & access	RBAC, users, groups, roles	`get users`, `create user`, `apply role`
Accounts	Cloud and VCS integrations	`create account`, `create github-account`, `get accounts`

Run raptor <command> --help on any item above for full flags, examples, and aliases.

Two words that are easy to conflate: an artifact is the named CI integration a resource deploys from, and a build is one image or zip registered under it. create artifact and delete artifact act on the first; get builds, promote build and delete build act on the second. To find which artifact a resource deploys from, read the ARTIFACT column of raptor get resources -p PROJECT -o wide.

Validate before you release

raptor plan is the pre-release gate. Run it against a project and environment before you create a release, and read the output before you apply anything:

raptor plan -p PROJECT -e ENV

A plan runs pre-flight validation locally (expression resolution, dependency cycles, input wiring) and then requests a Terraform plan from the Control Plane, so you see both the configuration errors and the infrastructure changes the release would make.

Useful flags:

Flag	What it does
`--skip-server-plan`	Runs local validation only and skips the server-side Terraform plan. Fastest way to check a blueprint edit.
`--target KIND/NAME`	Plans only the named resources, for example `--target service/api`. Repeat the flag for more than one.
`--force`	Produces a plan even when Facets detects no pending changes.
`--allow-destroy`	Allows resource destruction in the plan.
`-m, --message`	Attaches a comment to the plan.

# Local validation only
raptor plan -p myproject -e dev --skip-server-plan

# Plan a single service, allowing resource destruction
raptor plan -p myproject -e dev --target service/api --allow-destroy

CI/CD workflows

The most common use of the CLI is in a CI pipeline that builds a container image and registers it as a Facets artifact. See Integrating with CI pipelines for the end-to-end build, push, register, deploy flow.

Self-update

Raptor can update itself in place:

raptor upgrade

AI agent skills

Raptor embeds its own CLI skills, which teach an AI coding agent how to drive raptor itself. They come from the binary, so they match the version you have installed. Install them into a supported agent's config directory:

raptor install-skills --agent claude    # also: gemini, codex
raptor install-skills --upgrade         # refresh previously installed skills

Skills install globally by default. Pass --path <dir> to install them at the project level instead.

These are separate from your organisation's skill catalogue, which lives on the control plane and is installed by praxis login. Most people want both: see Build with an AI agent.