Radoslaw Fabisiak

@gokuljd Congrats on the launch. Closing the loop between chat and live deployment is a powerful shift, especially if projects can keep evolving instead of being one-shot generations. I’m curious how you handle trust boundaries when the model can modify backend logic and database schema through MCP. If a prompt or imported snippet contains adversarial instructions, what safeguards prevent...

Congrats on the launch @ron_shahar. Designing a backend specifically for AI agents instead of forcing them through human-first APIs is a meaningful architectural shift, especially with Skills and built-in connectors abstracting away auth and setup. Curious how you’re thinking about trust boundaries between agents and Skills. If an agent misinterprets intent or receives adversarial input, what...

Congrats on the launch. Correlating logs, metrics, traces, and deployment history into a single investigation flow is a hard problem, and the "investigation takes longer than the fix" insight really resonates. I’m curious how you handle adversarial or malformed log input when it’s fed into the agent’s reasoning context. Do you isolate untrusted telemetry to prevent it from influencing root...