Track skipped Node compatibility tests from CI stabilization

[sonukapoor]

PR #74 restored the CI baseline by skipping a set of Node compatibility tests that were already failing or unstable in the native test-only job. The skips should be treated as follow-up runtime compatibility work, not as permanent ignore rules.

The failures seen during CI stabilization grouped into a few areas:

• HTTP/2 stream and lifecycle behavior:

  • parallel/test-http2-client-jsstream-destroy.js
  • parallel/test-http2-forget-closed-streams.js
  • parallel/test-http2-pipe.js

• TLS crashes / lifecycle behavior:

  • parallel/test-tls-alert-handling.js
  • parallel/test-tls-close-notify.js
  • parallel/test-tls-connect-abort-controller.js

• Domain abort / uncaught exception behavior:

  • parallel/test-domain-no-error-handler-abort-on-uncaught-0.js
  • parallel/test-domain-no-error-handler-abort-on-uncaught-1.js
  • parallel/test-domain-no-error-handler-abort-on-uncaught-2.js
  • parallel/test-domain-no-error-handler-abort-on-uncaught-3.js
  • parallel/test-domain-no-error-handler-abort-on-uncaught-4.js
  • parallel/test-domain-no-error-handler-abort-on-uncaught-5.js
  • parallel/test-domain-no-error-handler-abort-on-uncaught-6.js
  • parallel/test-domain-no-error-handler-abort-on-uncaught-7.js
  • parallel/test-domain-no-error-handler-abort-on-uncaught-8.js
  • parallel/test-domain-no-error-handler-abort-on-uncaught-9.js
  • parallel/test-domain-throw-error-then-throw-from-uncaught-exception-handler.js
  • parallel/test-domain-with-abort-on-uncaught-exception.js

• HTTP keep-alive / timeout behavior:

  • parallel/test-http-server-headers-timeout-keepalive.js
  • parallel/test-http-server-request-timeout-keepalive.js

• Abort tests that currently hang:

  • abort/test-http-parser-consume.js
  • abort/test-zlib-invalid-internals-usage.js

• Linux strace/OpenSSL behavior:

  • parallel/test-strace-openat-openssl.js

Suggested order:

1. Start with TLS SIGSEGV cases because crashes are the highest-risk failures.
2. Then fix HTTP/2 stream lifecycle cases.
3. Then address Domain abort semantics and the hanging abort tests.
4. Finally revisit the Linux strace/OpenSSL expectation and HTTP keep-alive timeout cases.

Useful verification command:

NODE_TEST_RUNNER=build-edge/edge ./test/nodejs_test_harness --category=node:http2,node:tls,node:domain,node:http,node:zlib -j 4

Each cluster should probably be fixed in a separate PR so failures stay easy to review and bisect.