Where does the bug appear (feature/product)?
Cursor IDE
Describe the Bug
I am using a SKILL which describes the coding process. This SKILL must have spawn specific subagents described in dedicated files for subtasks e.g. coding, testing, reviewing. After I updated Cursor to v3.0+ it stopped working.
I see the chat is trying to start new agent but it does not change anything, does not spend tokens and etc. Nothing happens for hours.
Steps to Reproduce
- make a SKILL which tells to Cursor it should spawn subagents for coding, reviewing, testing
- make a command with a simple instruction like "Use the
myskill skill at skills/myskill/SKILL.md. Do not perform any work outside that skill.`
- run it as the prompt
/mycommand implement...
- wait until it starts spawning subagents
Expected Behavior
Everything works, I can open the subagent’s information and see what it is doing. Code updates, tokens spends
Screenshots / Screen Recordings
Operating System
MacOS
Version Information
Version: 3.1.17
VSCode Version: 1.105.1
Commit: fce1e9ab7844f9ea35793da01e634aa7e50bce90
Date: 2026-04-19T19:33:58.189Z
Layout: editor
Build Type: Stable
Release Track: Default
Electron: 39.8.1
Chromium: 142.0.7444.265
Node.js: 22.22.1
V8: 14.2.231.22-electron.0
OS: Darwin arm64 25.4.0
Does this stop you from using Cursor
Sometimes - I can sometimes use Cursor
For the context, I am working on that skill since Cursor v2+ and it was working well. Now it does not perform tasks at all
This is being actively tracked by our engineering team as a high-priority item. No fix has shipped yet, but it’s under active investigation.
One diagnostic that could help: when the subagent gets stuck on “Starting up,” open the Developer Console (Help > Toggle Developer Tools > Console tab) and check for any red error messages at the time the subagent launches. A screenshot of those errors would help the team narrow down what’s failing client-side.
Also, a quick question about your orchestration workflow: does your SKILL have the main agent spawn subagents directly, or does it use a multi-layer pattern where a coordinator subagent then spawns additional worker subagents? The answer affects which fix path applies.
Hey Mohit, I prepared a few screenshots. It looks different for now, at least 3/4 tasks are complete with the orchestration but the last one is complete stuck the entire process.
Also, a quick question about your orchestration workflow: does your SKILL have the main agent spawn subagents directly, or does it use a multi-layer pattern where a coordinator subagent then spawns additional worker subagents? The answer affects which fix path applies.
My orchestration setup is a single-level pattern: the main agent (the one that loads and follows the orchestration skill) spawns subagents directly for each role/stage (Planner, Planner-Validator, Implementer, Reviewer, Tester, Verifier).
It does not describe a multi-layer pattern in which a coordinator subagent is the only one that spawns additional worker subagents. Multiple subagents in a row (e.g. Planner then Planner-Validator, or Implementer then Reviewer) are still all spawned from the same orchestrating agent, not from a dedicated coordinator subagent.
For the context, what am I doing here?
I have an orchestration SKILL which analyzes the incoming task (list of tasks) and spawns subagents for building a plan, asking clarifying questions, writing the code, reviewing the code, covering the updates with unit tests, running the quality gates and verifying that the task is complete.
In the example I have a workspace of 4 repositories and the code writing subagent for one of them stuck. All the rest have done their jobs well this time.
The screenshots are below:
