In the world of GRC, we are at a threshold where the rapid use of AI Agents demands a shift left approach to identifying, implementing and validating controls.

That makes sure the controls are valid through out the lifecycle of Agentic workflow, change mgmt is implemented as the Agents evolve.

For the answer, we can look at the evolution that the world of Software Testing - Quality Engineering.

Today, the shift left approach of testing is not just advisable but a norm, to an extent that it might be the preferred and only way to ensure the features delivered are quality assured.

In the past, the GRC team had an infrequent look at the risks ,In the Agentic world though, the infrequent validation of risks and compliance playing catch up in a losing race.

The question here is - Whether there is a better way to build & enforce guardrails?

Qwen and the Kimi series are really good.

Sounds like the old school way of Wordpress. The reader was barely useful in the context as the feed is nowadays.

The middle ground is an easier way to keep track of subs while the focus is solely on writing and distribution / integration / analytics