mcpField
G
gb-setup

gb-setup

github.com/growthbookDevelopment25102 installsUpdated 4w ago
Repository

About gb-setup

gb-setup is an onboarding skill for GrowthBook Agent Skills users connecting to GrowthBook Cloud or a self-hosted instance. It validates credentials and the app origin, then stores configuration with restricted permissions.

Key features

  1. 01Detects existing configuration and masks secrets
  2. 02Validates API keys and HTTPS origins
  3. 03Checks credentials and the app origin
  4. 04Writes a permission-restricted config file
  5. 05Allows environment variables to override file values

Use cases

  1. 01Setting up GrowthBook Agent Skills for the first time
  2. 02Configuring API and UI origins for self-hosted GrowthBook
  3. 03Validating access credentials before saving configuration

Install

Copy the command and run it in a terminal.

$npx skills add https://github.com/growthbook/skills --skill gb-setup

After installing, type/gb-setupin Claude Code to confirm it loaded.

Details

From the project README

gb-setup

One-skill onboarding for GrowthBook Agent Skills. Walks the user through GB_API_KEY plus GB_API_URL and GB_APP_URL for self-hosted. Validates the credentials and trusted app origin, then writes ~/.config/growthbook/.env with chmod 600.

The API key is a Personal Access Token (PAT) tied to a GrowthBook user, so the API attributes any flags or experiments the write skills create to that user automatically — there's no separate owner identifier to configure.

gb-call reads this file when the corresponding environment variables aren't set, so the user gets a one-time config rather than editing their shell rc. Real environment variables always win over the file — useful for CI and one-off overrides.

Resolve the bundled helper for the current client before using it: under the Claude Code plugin it is ${CLAUDE_PLUGIN_ROOT}/scripts/gb-call; under a standalone Agent Skills install it is scripts/gb-call relative to this skill directory. Substitute that resolved command for <gb-call> below.

Workflow

  1. Detect current state. Check what's already configured. Don't ask the user for values they already have unless they want to change them.

    test -f ~/.config/growthbook/.env && echo "exists" || echo "missing"
    

    If the file exists, read it with the Read tool, parse the KEY=value lines, and surface a masked summary:

    I see existing config at ~/.config/growthbook/.env:
      GB_API_KEY = gb_pat_****wxyz   (last 4 shown)
      GB_API_URL = (not set — defaults to https://api.growthbook.io)
      GB_APP_URL = (not set — defaults to https://app.growthbook.io because the API URL uses GrowthBook Cloud)
    
    Want to keep these, update one, or start fresh?
    

    Describe an unset GB_APP_URL based on the effective API origin: it defaults to https://app.growthbook.io when the API origin is GrowthBook Cloud, but is required when a custom self-hosted GB_API_URL is configured.

    Also note what's in process.env — if the user has shell exports, those will override the file. Surface that ("GB_API_KEY is also set in your shell environment; the file value won't be used unless you unset the shell var.")

  2. Collect GB_API_KEY (required). If keeping the existing value, skip. Otherwise, show the transcript-exposure notice first so the user can make an informed choice:

    Before you paste your key: anything you type into this chat may be stored by your agent client and is sent to your configured model provider as part of the conversation. The skill will mask the key in its replies, but the value you paste cannot be masked retroactively. Check your client's storage and retention settings if this is a concern.

    Recommendation: generate a fresh PAT for this plugin rather than reusing your personal admin token. That way you can revoke it independently if anything goes wrong, without affecting your other API access.

    Then ask:

    Paste a Personal Access Token (PAT) or Secret Key. Get one at:

    PATs start with gb_pat_; Secret Keys with secret_. Either works.

    Once captured, never echo the value back in any later step — mask all but the last 4 characters.

  3. Ask about self-hosted (optional GB_API_URL and GB_APP_URL).

    Are you using GrowthBook Cloud (api.growthbook.io) or a self-hosted instance?

    • Cloud → skip, leaving both URLs unset (gb-call defaults to https://api.growthbook.io for API calls and https://app.growthbook.io for UI links).
    • Self-hosted → prompt separately for:
      • The API origin stored as GB_API_URL (e.g. https://api.acme-internal.com).
      • The user-facing GrowthBook UI origin stored as GB_APP_URL (e.g. https://growthbook.acme-internal.com).

    Do not derive one origin from the other. Self-hosted deployments may use unrelated hostnames or serve the API and UI from the same origin.

    Validate each URL's shape before accepting:

    • Must parse as a URL (i.e. new URL(value) does not throw).
    • Scheme must be https:// (refuse http:// — auth headers over cleartext leak credentials; refuse file://, data://, etc.).
    • Hostname must be non-empty.
    • No username or password.
    • No path, query, or hash — new URL(value).pathname must be / or empty, with empty search and hash. A value like https://api.acme.com/v1 would silently mis-route requests or links; reject with a message that asks for just the origin.
    • Strip a single trailing slash before storing.

    Refuse any URL that fails these checks and loop back to ask again; don't silently coerce. Silent fix-ups train users to trust that the system "just works" when the value is sometimes wrong in ways the system can't fix.

  4. Validate against the live API. Set the collected values as ad-hoc env vars and call a lightweight authenticated endpoint:

    GB_API_KEY='<value>' \
    GB_API_URL='<cloud default or self-hosted API origin>' \
    GB_APP_URL='<cloud default or self-hosted app origin>' \
      <gb-call> GET /api/v1/projects
    

    Interpret the result:

    • 2xx → credentials work. Move on.
    • 401 / 403 → "Authentication failed. The API key is invalid, expired, or has been revoked. Generate a fresh one at <url-from-step-2> and try again." Loop back to step 2.
    • 404 on a growthbook.io host → "Got a 404 from api.growthbook.io. If you're actually on self-hosted, set GB_API_URL to your API base URL." Loop back to step 3.
    • Network error / DNS → "Couldn't reach <host>. Check the URL and your network." Surface the raw error.

    Don't proceed past validation. If the user is stuck, halt and let them debug — don't write a broken config.

    Then verify that the trusted app origin resolves exactly as expected:

    GB_API_URL='<cloud default or self-hosted API origin>' \
    GB_APP_URL='<cloud default or self-hosted app origin>' \
      <gb-call> app-origin
    

    The output must equal the expected UI origin with no trailing slash. If it fails, return to step 3.

  5. Write ~/.config/growthbook/.env — order matters for security.

    Create the directory and lock it down first:

    mkdir -p ~/.config/growthbook
    chmod 700 ~/.config/growthbook
    

    The directory's 0700 (owner-only) permission is what protects the file from being readable by other users on the system during the next step. Writing the file first and then chmod-ing it leaves a window where the file inherits the user's umask (typically 0022 → mode 0644, world-readable). Locking the directory closes that window — files inside a 0700 directory aren't reachable by other users regardless of file mode.

    Then write with the client's file-writing tool. Format — one KEY=value per line, no quoting (our values never contain spaces, =, or newlines):

    GB_API_KEY=<value>
    GB_API_URL=<value or omit line>
    GB_APP_URL=<value or omit line>
    

    Then set the file mode as belt-and-suspenders:

    chmod 600 ~/.config/growthbook/.env
    

    0600 means owner-read/write only. With the directory at 0700 and the file at 0600, the secret is unreachable to any other user on the system.

  6. Report. Tell the user:

    • Where the file is (~/.config/growthbook/.env).
    • What's in it (mask the API key — show only last 4).
    • The trusted UI origin that gb-call app-origin returns.
    • That env vars take precedence over the file (so CI / one-off overrides keep working).
    • That re-running the gb-setup skill updates the file.
    • The next thing they probably want: the feature-flags skill (flag-search workflow) to see their flags, or the experiments skill (experiment-brainstorm workflow) to look at past results.

Guardrails

  • Never echo the API key back in plain text. Always mask except for the last 4 characters. The skill's output ends up in the user's terminal and transcript.
  • Surface the transcript-exposure risk before the user pastes a key. The value typed into chat may be stored by the agent client and is sent to the configured model provider; the skill cannot retroactively redact it. Always recommend a freshly-scoped PAT over reusing an admin token. This isn't paranoia — it's the right way to handle a workflow that requires a user to paste a secret into a conversational interface.
  • Revocation guidance is a real fix, not a footnote. If a key was ever exposed, the only effective remediation is to revoke and rotate it at <GB_APP_URL>/account/personal-access-tokens (or the cloud default). Surface this whenever the user expresses concern about an old or shared key.
  • Directory at 0700 before file write; file at 0600 after. The order matters. The Write tool inherits the user's umask, which on most systems creates files at mode 0644 (world-readable). Locking the directory down first means even the brief window between Write and chmod is not reachable by other users. Then chmod the file as defense in depth. Both steps are non-optional; skipping either turns the PAT into a leak.
  • Env vars take precedence over the file per variable. If a value is set in process.env, gb-call ignores that variable's file value. Surface this if both are configured so the user understands what's actually in effect.
  • Validate against a real endpoint, not just shape. A token that "looks right" but doesn't work is worse than no token — the user thinks they're set up. GET /api/v1/projects is lightweight, requires auth, and works on every plan tier.
  • No owner identifier to collect. The PAT is tied to a GrowthBook user, so the write skills (flag-create, experiment-launch) let the API attribute new flags and experiments to the token's user automatically. Don't prompt for an email or userId during setup.
  • Don't strip leading whitespace from line values, but strip trailing. A PAT shouldn't have whitespace at all; if the user pasted one with a trailing newline from a clipboard, strip it. Leading whitespace is unusual enough that surfacing rather than silently fixing is safer.
  • Don't write a partial file. If validation fails, halt without touching ~/.config/growthbook/.env. The user re-runs setup; we never overwrite a working config with a broken one.
  • Strip trailing slashes from both URLs. gb-call normalizes GB_APP_URL too, but storing clean origins avoids inconsistent output and double-slash bugs.
  • Never infer GB_APP_URL from GB_API_URL. The api. → app. convention is not guaranteed for self-hosted deployments. Ask for the app origin explicitly.
  • Don't propose env vars we don't actually use. Only GB_API_KEY, GB_API_URL, and GB_APP_URL. Anything else is scope creep and will confuse future contributors.

File format reference

~/.config/growthbook/.env:

GB_API_KEY=gb_pat_abc123def456
GB_API_URL=https://api.your-self-hosted.com
GB_APP_URL=https://growthbook.your-self-hosted.com

Lines starting with # are comments. Blank lines are ignored. No quoting; values are taken verbatim from the first = to end of line.

Endpoints used

  • GET /api/v1/projects — validation only. Lightweight, requires auth, available on every tier.

Handoffs

  • The feature-flags skill (flag-search workflow) — natural first call after setup completes (no required inputs, exercises the new config).
  • The experiments skill (experiment-brainstorm workflow) — if the org already has stopped experiments, surfaces them immediately.
  • Any skill that emits a config-related error points back here.

Comments

Markdown supported