fix: handle spawn ENOENT errors and prevent server crash

[Priyanshu-u07]

Summary

This PR fixes a crash in the FaaS server that happens when the metacall binary is missing from the system's PATH. It handles the ENOENT error when starting child processes and prevents the server from crashing repeatedly during startup when trying to reload failed deployments.

Related issue

Fixes #74

Type of change

• Bug fix
• New feature
• Documentation
• Chore / CI
• Breaking change

How to test

1. Simulate missing dependency: Start the FaaS server in an environment where metacall is missing.
   • Windows (PowerShell):
     $env:PATH = "C:\Windows\system32;C:\Program Files\nodejs;C:\Windows\System32\WindowsPowerShell\v1.0\"; npm start
   • Linux: PATH="/usr/bin:/bin" npm start
2. Trigger the bug: In a separate terminal, attempt to deploy any application:
   metacall-deploy --dev
3. Verify survival: Observe that the server logs a warning (spawn metacall ENOENT) but remains online.
4. Verify recovery: Restart the server. Observe that it successfully boots by skipping the broken app in the cache instead of crashing.

Checklist

• I have read the contributing guidelines
• I added tests that prove my fix is effective or that my feature works
• I updated documentation if necessary

Notes for reviewers

• Fix Strategy: We added a proc.on('error') listener in deploy.ts to handle spawn failures gracefully without crashing.
• Data Safety: In autoDeploy.ts, the implementation now logs and skips failures instead of deleting the cached app folder. This prevents startup crash loops while keeping user data safe during temporary environment issues.

Release notes

Fixed a fatal server crash and startup loop occurring when the MetaCall binary is missing from the environment PATH.