Sandboxes Shape What You Are Capable of Automating

Out of all the tasks you accomplished today, how many of them could be automated? It's a difficult question to answer.

As a developer, what individual tasks, operations, and actions did you perform against GitHub, Jira, or Claude today? Which actions did you take against your calendar, Slack, or the three Zoom meetings you attended? How much of this can you automate? What is even possible? What isn't?

Even with these open questions, we enjoy dreaming about the possibilities when it comes to agentic automation—even before we know what we're actually capable of.

Let's pause for a moment, gather the signals across our busy days, and explore how we use the OpenAPI and JSON Schema specifications to curate API sandboxes. We can use their examples and tags to define what we accomplish each day and what should be automated—organizing all of the operations from APIs we depend on across GitHub, Jira, Claude, Google Calendar, Slack, Zoom, and other services into a single place. Then, using these OpenAPI artifacts, we can mock all of the APIs we use and leverage the open-source Bruno API client to demonstrate what we're capable of today in our sandboxes via our IDEs, copilots, and agents.