failCmd doesn't get invoked when publishCmd returns non-zero error code

semantic-release-errorlog.log
I'm trying to create a workflow where I should be able to notify on success and failure of publishCmd, but failCmd doesn't get invoked at all when publishCmd returns non-zero error code. Instead the semantic-release crashes with error stack trace.

{
  "ci": false,
  "branches": [
    "test",
    {
      "name": "beta",
      "prerelease": true
    },
    {
      "name": "alpha",
      "prerelease": true
    }
  ],
  "plugins": [
    "@semantic-release/commit-analyzer",
    "@semantic-release/release-notes-generator",
    [
      "@semantic-release/changelog",
      {
        "changelogFile": "docs/CHANGELOG.md"
      }
    ],
    [
      "@semantic-release/exec",
      {
        "failCmd": "./fail.py",
        "successCmd": "./success.py",
        "publishCmd": "./deploy.sh \"${nextRelease.version}\"",
        "verifyConditionsCmd": "./verify.py"
      }
    ],
    [
      "@semantic-release/git",
      {
        "assets": [
          "docs/CHANGELOG.md"
        ],
        "message": "chore(release): ${nextRelease.version} [skip ci]\n\n${nextRelease.notes}"
      }
    ]
  ]
}

Reproduce the workflow using the shown publishCmd, failCmd, successCmd, and verifyConditionsCmd configuration, then inspect the semantic-release/exec handling around publishCmd errors. Compare the observed error log with the expected failure notification behavior; done means failCmd is invoked when publishCmd exits non-zero without the release crashing first.