Electron: Is it safe to expose ipcRenderer methods with only a startsWith if statement “protection”?

I have an electron application and I’m using contextBridge.exposeInMainWorld to expose some ipcRenderer methods such as .on , .removeListener and .invoke, these methods are “protected” by having an if statement that checks if the “channel” passed is valid.

Using the following code:

const validateIPC = (channel) => {
  if (!channel || !channel.startsWith("myapp:")) 
    return false;
 
  return true;
};

contextBridge.exposeInMainWorld("electron", {
ipcRenderer: {
    on(channel, listener) {
      if (validateIPC(channel)) {
        ipcRenderer.on(channel, (evt, message) => {
          listener(evt, message);
        });
      }
    },
    removeListener(channel, listener) {
      if (validateIPC(channel)) {
        ipcRenderer.removeListener(channel, (evt, message) => {
          listener(evt, message);
        });
      }
    },
    invoke(channel, data) {
      if (validateIPC(channel)) {
        return ipcRenderer.invoke(channel, data);
      }
    },
  },
};

Is it safe to only check if the channel starts with some random string? I saw this on the source code of vscode available on github, however I also read some solutions using an array of strings to validate the channels. If it is safe why can’t I just expose it without checking if the channel startsWith some random name? Are there any default channel that can’t be exposed?