Solutions · Production readiness
Know the stack is ready before you ship
Tests check your code. They don't check that production's webhook exists and verifies signatures, that the auth provider allows your production URL, or that a test key isn't in production. Ahena checks the external stack your code depends on, and can block a deploy that isn't ready.
What breaks on launch day
The failures that tests and code review don't catch:
- A webhook endpoint that's missing, disabled, or accepts unsigned requests.
- A production redirect or site URL the auth provider doesn't allow.
- A test-mode key in production, or a production secret that was never stored.
- A sending domain whose DNS records were never verified.
- A credential that expired or lost a permission.
What Ahena checks
Doctor
ahena doctorruns each connected provider's checks against its real state (credentials and permissions, configuration, webhooks, domains, CORS, migrations and more, per provider) plus environment separation and required secrets. Every finding says what's wrong and how to fix it.A deploy gate
ahena deploy-checkruns Doctor for one environment, plus drift againstahena.lock, and exits 1 when it isn't ready to deploy, with each blocker labelled.In CI
ahena ci initwrites a GitHub Actions workflow that runs the check with a read-only, project-scoped token.Fixes through plans
Findings Ahena can fix become a plan you approve (
ahena doctor --fix); the rest list the exact manual steps.
What deploy-check reports
| Blocker | Typical cause |
|---|---|
| missing production secret | A secret the app needs isn't stored |
| invalid provider credentials | A key was revoked, expired or lacks permissions |
| configuration drift | A high-severity value changed outside Ahena |
| broken webhook | Endpoint missing, disabled, or accepting unsigned requests |
| development credential in production | A test-mode key configured in production |
| missing OAuth callback | Production redirect or site URL not allowed |
| missing provider | The app needs a provider that isn't connected |
Before a release
$ ahena doctor -e production$ ahena deploy-check --environment production # exit code 1 if it isn't ready$ ahena ci init # the same check on every push
Good to know
- Ahena checks provider configuration, not your application: it doesn't test your code or guarantee it has no bugs.
- Checks vary by provider; each integration page lists its Doctor checks.
- Warnings don't fail
deploy-checkunless you pass--strict.
Questions
How do I verify provider configuration before deployment?
Run ahena doctor -e production to check each connected provider's real configuration, and add ahena deploy-check --environment production to CI so a deploy fails when a required secret, webhook, redirect URL or credential isn't right.