← sources

CC Agent Opus speaking with another internal model. Has anyone noticed this?

source: reddit (reddit:ClaudeAI)published 2026-10-27 21:48ingested 2026-10-28 02:35

retained text

Has anyone else ever noticed your CC Agent CLI talking to something else when it’s working? 

I understand it’s a reason/react pattern, also to keep us in the loop. But there are a few times I’ve watched it seemingly get stuck, then reason or discuss with something, and then aha, it continues. 

There’s once I watched it go through that same scenario. It discussed its point on some idea or concept, asking how could something even be possible due to X. Then suddenly,  text I’ve never seen before in a different color responded to it in a corrective/explanatory manner. And again, aha, and continued. 

From my understanding, our own sandboxed iteration is what we get. But it used to make me wonder if there was a more intelligent model not allowed for public use, that could help ours in occasion. Lmfao

document fetched 2026-10-28 00:21 · canonical https://reddit.com/r/ClaudeAI/comments/1xh5wfa/cc_agent_opus_speaking_with_another_internal