Multiple agents for the same prompt

[ahtisamafzal]

I'm exploring whether it's possible to invoke multiple agents for the same prompt — similar to a Claude-flow swarm or hive-mind setup — where we simultaneously query several models like:

• Claude
• GLM 4.6
• Codex GPT
• OpenAI GPT

The goal would be to compare their outputs side by side and select the best response based on quality, relevance, or style.

Codex already supports this kind of multi-agent generation, allowing up to four variations of the same prompt. This has been especially useful for UI-focused workflows, where visual and interaction design benefits from diverse interpretations.

Cursor 2.0 recently expanded on this by introducing model comparison features, with all outputs tracked in a GitHub work tree — making it easy to audit, version, and refine collaboratively.

Is this kind of multi-agent orchestration something you could support or integrate.