These are platform capabilities, so they behave the same way in every product rather than being configured separately in each.
Treepie does not integrate its own products with each other — there is nothing to integrate. Integrations exist for the systems it does not replace: banks, tax authorities, calendar and email, and whatever industry-specific tool a company genuinely needs.
The API reads and writes the same record the products use, with the same permission model applied. Webhooks fire on record changes rather than per-product events, so a subscriber does not need to know which product made the change.
We are cautious about connectors that recreate a product Treepie already has. An integration that syncs a second CRM into Flow gives you two customer records and a sync — which is the exact problem the platform exists to remove. If you need both, keeping them deliberately separate is more honest than pretending a sync makes them one.
API access is scoped the same way a person is: a key can only do what the role attached to it can do. Keys are visible in administration with their last-used time, because the most common credential problem is not a leak, it is a key nobody remembers creating still working two years later.
Webhooks fire on record changes with a stable event id, and consumers are expected to deduplicate on it. Delivery is retried with backoff. We would rather send a duplicate you can discard than drop an event you needed.
The API is included from the Standard plan. Starter covers records, permissions and support but not programmatic access.