The Agent Stack Is Quietly Consolidating Around tool-first agents

For most of the last year, the conversation around structured outputs has been louder than the evidence. That is starting to change.

What DeepSeek actually shipped with Claude 4.5 Sonnet is less a single capability and more a cluster of small, compounding improvements — the kind that only show up when you put a real workflow on top.

Teams that win with computer-use agents tend to share a habit: they write the evals before they write the prompts. Everything else follows from that.

Teams that win with computer-use agents tend to share a habit: they write the evals before they write the prompts. Everything else follows from that.

The skeptical read is that we are watching a feature, not a platform. The optimistic read is that code execution is exactly the kind of feature that becomes a platform when nobody is paying attention.

The skeptical read is that we are watching a feature, not a platform. The optimistic read is that agent mode is exactly the kind of feature that becomes a platform when nobody is paying attention.

Inside Databricks, the rollout looked less like a moonshot and more like a slow migration. A pilot, a champion, a quiet expansion, a budget line.

None of this guarantees a clean story. OpenAI could ship a model next month that rearranges the assumptions in this piece. But the direction of travel, for now, is clear enough to plan around.