RevenueCat CLI

rc manages your entire RevenueCat integration from the terminal: customers, subscriptions, products, paywalls, and store credentials. It's built to be driven equally well by you and by AI coding agents.

• Developers who'd rather stay in the terminal than click through the dashboard
• CI pipelines that need stable --json, --no-input, and predictable exit codes
• AI agents with full schema discovery, one-command setup, and installable skills

rc auth login
rc customers show cus_abc
rc charts show mrr

Set up in one command

From your app's directory, with no install and no dashboard clicking:

npx @revenuecat/cli setup

An AI agent walks the whole RevenueCat setup while you approve each step, and emits a prompt your own coding agent can run non-interactively. Already have a project? Point it straight at a store:

rc setup apple <app-id>      # App Store Connect: create + upload keys, vendor number
rc setup google <app-id>     # Google Play: bootstrap + upload the service-account credential

<app-id> is the RevenueCat app ID (like app_abc), not a bundle ID or a store app ID. Installed via Homebrew or npm, drop the npx @revenuecat/cli and just run rc setup. Details on the store flows are in Store credential setup below.

Install

Homebrew (macOS/Linux):

brew install RevenueCat/tap/rc

npm / npx (no Homebrew required; ideal for CI, agent sandboxes, and React Native projects). The package ships the same native Go binary; Node only dispatches:

npx @revenuecat/cli --help
# or install it
npm install -g @revenuecat/cli

Or download a binary directly from the releases page.

Homebrew and npm install the command as both rc and revenuecat (same binary); use whichever you prefer. Examples throughout use rc.

Every command also runs without installing, via npx @revenuecat/cli <command>, handy for CI and agent sandboxes.

Quick start

The fastest way in is rc setup above. Prefer to do it by hand:

# 1. Log in (browser OAuth or paste an API key)
rc auth login

# Or create a new account and log in without a browser
rc auth signup

# 2. Set a default project (or choose "Ask me every time" for multi-project workflows)
rc projects use

# Set up store-side credentials (App Store Connect / Google Play)
rc setup apple app_abc

# Look up a customer
rc customers show cus_abc

# See your MRR in an interactive chart
rc charts show mrr

What you can do

Area	Commands
Setup	setup (guided) · setup apple · setup google
Customers	show · list · grant · revoke · transfer · aliases · attributes · simulate-purchase
Subscriptions	show · cancel · extend · refund · transactions
Entitlements	list · show · create · update · attach · detach
Offerings	list · show · verify · preview · create · update · set-current · archive
Packages	list (across all offerings) · show · products · create · update · delete · attach · detach
Products	list · show · create · archive · restore · delete · store sync
Paywalls	list · show · create · generate (AI) · edit (AI) · rewind · publish · unpublish · attach · detach · delete
Rico (AI assistant)	rico (streaming chat, tool approvals) · conversations list/show/delete · feedback
Charts & metrics	Interactive bar/line charts · daily/weekly/monthly/quarterly/yearly
Apps	list · show · create · update · delete · apple check
Audit log	rc audit with --limit and --since
Webhooks	list · show · create · update · delete

rc --help          # see everything
rc customers --help # see subcommands for any noun

Verify a Test Store setup

Agents can inspect the complete offering graph, see the payload delivered to an SDK, and create a real headless Test Store transaction without raw API calls:

rc offerings verify ofrng_default --json --no-input
rc offerings preview app_test --app-user-id demo-user --json --no-input
rc customers simulate-purchase \
  --app-id app_test --product premium_monthly --app-user-id demo-user \
  --yes --json --no-input

offerings verify includes packages, attached products and prices, matching entitlements, paywall publication state, and an issues array. offerings preview uses the app's public SDK key and exposes the v1 SDK response; a null paywall_components value indicates fallback content rather than a published dashboard paywall.

Product store-state plans (experimental)

Manage your App Store / Google Play catalog as a reviewable plan: describe the desired state, review the computed diff, then apply it.

rc products store sync app_abc                       # interactive: review, then apply
rc products store sync app_abc --file catalog.csv    # from a CSV or JSON adapter

For CI and agents, separate review from mutation. plan returns a persisted plan ID that later commands apply by reference, so nothing local needs to survive between steps:

plan_id=$(rc products store plan app_abc --file catalog.csv --json --format '.data.id')
rc products store show "$plan_id" --json             # review the diff
rc products store apply "$plan_id" --yes             # or: discard "$plan_id"

CSV/JSON input is optional (RC_STORE_STATE_FILE can replace --file); the backend, not the local filesystem, is the durable handoff. This bulk import is experimental and may not be available on every account yet.

Store credential setup (experimental)

Set up the store-side credentials RevenueCat needs, without manual key downloads or console clicking. Both run locally: your Apple/Google sign-in stays between you and Apple/Google, and the CLI uploads only the resulting API keys (Apple) or service-account credential (Google) to RevenueCat so it can talk to the store on your behalf.

Apple (App Store Connect) creates the missing In-App Purchase and App Store Connect API keys, and uploads them to an existing RevenueCat App Store app:

rc setup apple app_abc

Google (Google Play) signs in locally, bootstraps the service-account credential, grants package-scoped Play access, and uploads it to RevenueCat:

rc setup google app_xyz

Preview Apple sign-in, two-factor authentication, team selection, and read-only key-management access without creating keys or changing RevenueCat:

rc apps apple check app_abc

The command supports trusted-device and SMS verification. Your Apple Account credentials are sent directly to Apple; they are never sent to RevenueCat or stored by rc. Newly created private keys are uploaded directly to RevenueCat and are never saved locally or printed. Interactive password entry is masked. For scripts, prefer RC_APPLE_PASSWORD over --apple-password to avoid shell history and process-list exposure. Pass --vendor-number if RevenueCat should also receive it. Small Business Program dates are intentionally deferred.

Profiles

Keep separate credentials for staging and production:

rc auth login --profile staging
rc auth login --profile prod

rc --profile staging customers list
rc --profile prod   customers list

Agentic signup

An agent can create an account and receive renewable OAuth credentials without opening a browser:

rc auth signup \
  --email dev@example.com \
  --name "Example Developer" \
  --generate-password \
  --save-password \
  --accept-terms \
  --no-input --json

Key points:

• Consent is explicit. Pass --accept-terms only after the user authorizes accepting the Terms of Service and Privacy Policy; add --marketing-emails only on a separate opt-in. --name is a personal display name, not a company.
• Passwords stay out of the shell. --generate-password --save-password creates a one-time password in memory and stores it in the macOS login Keychain without printing it. Prefer RC_PASSWORD over --password, which can leak via shell history and process listings.
• Verify the result. Check that account_created, authenticated, and password_saved_to_keychain are true. Tokens are saved to the active mode-0600 profile, same as rc auth login.

The returned JSON points the agent to verify the email and use the AI Toolkit (below) to configure the project. For a manual start: rc projects create --name "My App" --use.

Scripting and agents

Every command supports --json for stable, machine-readable output:

rc customers list --json | jq '.data.items[].id'
rc customers show cus_abc --json
RC_API_KEY=sk_... rc entitlements list --json --no-input

Discover the full surface programmatically:

rc commands --json       # full command tree with capabilities
rc schema <cmd>          # flags, args, and examples for any command

Hit any API endpoint not yet in the CLI surface:

rc api GET /projects/proj_abc/customers
rc api POST /projects/proj_abc/offerings --body '{"lookup_key":"sale"}'

AI Toolkit

RevenueCat's official AI Toolkit provides agent workflows for project setup, SDK integration, catalog management, and health checks. Install its skills through the standard Skills CLI:

rc skills install            # global; installs the core project-setup skills
rc skills install --project  # repository-local instead
rc skills install --all      # the full skill catalog

rc delegates to the Skills CLI (npx skills add RevenueCat/ai-toolkit) so the workflows update in one place rather than shipping a stale copy. Global installs run in an isolated temp dir, so no lock file or hidden directory lands in your repo. Re-run rc skills install to update, then reload your agent to pick up changes.

Skills don't run on install; the agent selects one when a request matches its description. You can also name one explicitly:

Use the create-revenuecat-project skill to make the app in this directory
RevenueCat Test Store-ready end to end, then report every production-store
stage separately.

Run rc skills prompts for more copy-ready starter prompts. Under the hood the toolkit discovers every command with rc commands --json and rc schema <cmd> --json, and drives typed primitives (rc products prices set, rc offerings set-current, rc apps keys) instead of raw API calls.

Global flags

Flag	Description
--json	Machine-readable output
--no-input	Fail rather than prompt (for scripts/CI)
--profile <name>	Config profile (precedence: flag → $RC_PROFILE → rc profiles use → "default")
--api-key <key>	Override profile key ($RC_API_KEY)
--yes, -y	Skip confirmation prompts
--all	Also show experimental (unreleased) commands in help
--no-color	Disable ANSI color (also respects $NO_COLOR)
--format <expr>	jq expression applied to --json output

Custom request headers

RC_HEADERS sends extra HTTP headers on every RevenueCat request (v2 API, Rico, and the Paywall AI editor), as newline-separated Name: Value pairs. Use it to route or tag traffic without baking anything into the binary:

RC_HEADERS=$'X-Some-Header: value' rc offerings list

One header per line; supplied headers override the CLI's defaults.

Usage analytics

Every authenticated v2 API request carries a few headers so RevenueCat can see which commands are used and whether the CLI is driven by a human, an agent, or CI — derived from normal request logs, with no separate telemetry and no identifier of any kind:

• User-Agent — revenuecat-cli/<version> (<os> <arch>; go<goversion>).
• X-RC-CLI-Command — the command path only (e.g. paywalls.generate), never arguments, flag values, IDs, emails, or prompts.
• X-RC-CLI-Mode — interactive, agent (--json or --no-input), or ci (when $CI is set).

Set DO_NOT_TRACK=1 (per consoledonottrack.com) to drop the X-RC-CLI-* headers. Requests still go through, just unlabeled.

Exit codes

Code	Meaning
0	Success
1	Error
2	Bad usage
4	Authentication / authorization
5	Not found
6	Rate limited

Releasing

Releases are fully automated via GoReleaser. To cut a release:

1. Make sure main is green on CI.
2. Tag the commit:

   git tag v0.1.0 && git push --tags

3. GitHub Actions runs GoReleaser, which:
   • Cross-compiles for macOS, Linux, and Windows (amd64 + arm64)
   • Creates a GitHub release with binaries and checksums
   • Updates the homebrew-tap formula automatically

Prerequisites (one-time setup):

• Add a HOMEBREW_TAP_GITHUB_TOKEN secret to this repo with write access to RevenueCat/homebrew-tap

Contributing

See CONTRIBUTING.md for dev setup, testing, and how to send a PR.

For deeper conventions (architecture, agent-friendly patterns, where things live), see AGENTS.md.

License

MIT. See LICENSE.