(Continuously Updated) My Real-Time Review of Cursor Subagents

[Artemonim]

@Artemonim

• Senior AI (Main Agent) - GPT-5.2 XHigh
• Senior SWE - GPT-5.2 XHigh
• Middle SWE - Codex 5.2 XHigh
• Middle Coder - Gemini 3 Flash
• Junior Coder - Grok Code Fast
• Verifier - Gemini 3 Flash

As a result, I came up with the following set:

• Agent: GPT-5.2 XHigh
• Subagents:
• • DB: Opus 4.5 Thinking
• • Senior: GPT-5.2 XHigh
• • Middle+: Codex 5.2 XHigh
• • Middle: Gemini 3 Flash
• • Junior: Grok Code Fast
• • Verifier: Gemini 3 Flash
• Hidden Cursor Subagent:
• • explorer: Composer 1

I’ll tell you more in Agent Compass when I get around to updating it.

[liquefy]

I have got a question from your practical use, do you use any subagents to run in background? Have you figured out what happens if main agents spawns a subagent that runs in background and main agent ends? will subagent finish or terminated?

[Artemonim]

Techlead.mdc:

(...)
## Limitations and Recommendations

### Limitations
- CI is single-threaded, meaning that only one subagent can be running at a time.
(...)

[Artemonim]

On one of my current projects, I had to downgrade from Python 3.13 to Python 3.12. I tasked the Subagent with checking what was broken.

The Subagent spent 33 minutes ensuring the project passed CI correctly. Meanwhile, the Agent reported: “After downgrading, nothing critical broke.”

[Artemonim]

@Opus 4.5 Thinking

Order of called agents

1. Middle SWE — creating an Astro website with a demo portal
2. Verifier — checking Middle SWE results → OK
3. Middle+ SWE — setting up and testing the CI pipeline
4. Verifier — final verification → OK

Final status

CI is running. The site is being built. TODO.md and CI_TODO.md are complete.

[Artemonim]

1. Using Codex CLI, Codex-5.3 XHigh created a three-phase refactoring plan (DRY, SRP, YAGNI as soft GC).
2. I added two verifier agents to the project, based on Gemini 3 Flash and Codex-5.2 XHigh.
3. I tasked Codex-5.2 XHigh with fully implementing the plan, verifying itself with CI and G3F after each phase, and finally verifying the work with C5.2XH.

[Artemonim]

My jaw literally dropped when Gemini 3 Flash launched subagents of its own model for a task with parallelization capabilities

UPD: Not repeated on a similar task: he hung several languages ​​on the subagent and also called Composer, not himself.

[Artemonim]

My Subagent List v2

This article provides an overview of the specialized subagents. Each agent is designed for a specific role within the development workflow, from rapid prototyping to deep verification and problem-solving. 1. Verifier Model: gemini-3-flash-preview Description: Can be launched at the end of complex tasks to verify task completion. Mode: Read-only Agent Instructions: You are a subagent. Your task is to verify if the Senior AI that called you has actually completed the assigned tasks. Respon…

[Artemonim]

I’m working with GPT-5.4 Mini XHigh on a design document for a new small project. I gave it 15 points of feedback and asked it to discuss the project with Gemini 3.1 Pro and GPT-5.4 High. Before that, in the same chat, GPT-5.4 Mini XHigh conducted technical intelligence on OSS and third-party service APIs through GPT-5.4 Mini XHigh and Composer 2 subagents.

The screenshot shows only one request for pre-production documentation.
Composer 2 and GPT-5.1 Codex Max High are also used as verifiers.

The Subagents are beautiful