Custom MCP cannot be recognized by the agent

Where does the bug appear (feature/product)?
Background Agent (GitHub, Slack, Web, Linear)

Describe the Bug
Starting from May 1st, custom MCP services cannot be loaded in conversations on cursor.com/agents. Only MCPs from the default template can be recognized by the agent. Before May 1st, custom MCPs were recognized correctly.

Steps to Reproduce

  1. Configure a custom MCP.
  2. Ask the agent to list available MCPs in the conversation.
  3. The AI cannot find the custom MCP service.

Expected Behavior
Custom MCPs should be recognized and usable.

Screenshots / Screen Recordings

Operating System
Windows 10/11

Version Information
For AI issues: which model did you use?
Composer 2 and any other models.

For AI issues: add Request ID with privacy disabled
bc-22e10199-2a53-4fde-bded-af305a89fa58

Additional Information
MCPs I have tried:

{
  "mcpServers": {
    "mysql": {
      "command": "npx",
      "type": "stdio",
      "args": [
        "-y",
        "/mcp-mysql-server",
        "mysql://root:username@ip/database",
        "list,read,utility,create,update,delete,execute,ddl,procedure,transaction",
        "database_discovery,custom_queries,analysis"
      ]
    },
    "feishu-mcp": {
      "command": "npx",
      "args": ["-y", "feishu-mcp", "--stdio"],
      "env": {
        "FEISHU_APP_ID": "1",
        "FEISHU_APP_SECRET": "2"
      }
    }
  }
}

Does this stop you from using Cursor?
No - Cursor works, but with this issue.

Hi,

Thanks for posting on the Cursor forum!

To help ensure most of our users can participate in discussions, this forum is currently English only.

If you’d like, you can edit your post to translate it into English. Once it meets our guidelines, it will be automatically relisted.

Thanks for understanding!

Note: This is an automated detection system and sometimes makes mistakes. If your post is already in English, feel free to ignore this message, or just make a small edit and it will be reviewed again.

This is a known bug. Our team has identified the issue and is working on a fix.

In the meantime, if your workflow allows it, one workaround is to pass the necessary API credentials directly to the agent via the Secrets/Environment configuration and have the agent make API calls directly (instead of through the MCP).

We’ll update once the fix is deployed.

thank you for your reply