Add filesystem cleanup on workflow failure or cancellation

[Copilot]

Summary

• Addresses review feedback: workflow now cleans up filesystems created during the run if deployment fails, preventing dangling Lambda Cloud resources.

Changes

• Track whether filesystem was created (vs. reused from filesystem_id input) via filesystem_created output
• Add cleanup step with if: (failure() || cancelled()) that deletes filesystem only when created by this workflow
• Log deletion errors as GitHub warnings with error details for manual intervention

Testing

• Not run (reason: workflow change validated via YAML syntax check; requires Lambda Cloud API to integration test)

Review focus

• Conditional logic in cleanup step ensures existing user filesystems are never deleted
• Error handling captures both API failures and connection issues

Rollout / Notes

• Cleanup runs automatically on any workflow failure after filesystem creation step

---

💬 We'd love your input! Share your thoughts on Copilot coding agent in our 2 minute survey.