Hi, may I ask if code execution agent functionality is currently supported? Or is your team planning to develop such an agent feature in the future? For example, something similar to the code execution capability in SmolAgent (Secure code execution).
Hey @Hakstar
Thank you for reaching out and using Agno
While Agno doesn’t have a pre-built CodeAgent like SmolAgent, it’s super flexible and lets you build workflows where agents can execute code as part of their behavior. Everything is written in Python, so you can define exactly how your agents think, reason, and act—including running code when needed.
That said, you’d handle security yourself—for example, by validating inputs or setting up sandboxed environments if needed.
Pl refer: Agno Workflows – How Agents Run Code
Let me know if you have any more doubts
1 Like