Can AI Agents Be Real Team Members?

In software projects, AI agents can already help with coding, testing, debugging, and documentation. But they’re often still treated as simple assistants.

I’m exploring Sharkly.ai as an Asana alternative that treats agents as real teammates, where agents can take ownership of tasks and work alongside human team members.

Would you let an AI agent own a testing or development task in your project?