Fix Jupyter Nsight service proxies

[robobryce]

Summary

• install socat in every tutorial image that configures ACH_PORT_FORWARDS
• fail Jupyter startup immediately with an actionable error when port forwards are configured without socat
• restore the /nsys/ and /ncu/ named proxy paths for CUDA C++, CUDA Tile, nvmath-python, and stdpar

Root cause

The shared Jupyter entrypoint silently skipped creation of its localhost forwarders when socat was unavailable. jupyter-server-proxy then received browser requests normally but failed to connect to localhost:8080 and localhost:8081, returning HTTP 500. Direct Nsight service URLs continued to work because they bypassed those missing listeners.

This was reproduced on Brev node scipy-rtx-pro-6000-4576fe: Docker service connections succeeded while both local forwarding ports returned ECONNREFUSED.

Validation

• bash -n brev/entrypoint-jupyter-user.bash
• missing-socat container test confirms immediate exit 1 and actionable error
• docker buildx build --check for CUDA C++, CUDA Tile, nvmath-python, and generated stdpar Dockerfiles
• HPCCM stdpar generation includes socat
• docker compose config --quiet for all Brev Compose files
• repository pre-commit and manual hooks for changed files
• commit signature hook

[copy-pr-bot]

This pull request requires additional validation before any workflows can run on NVIDIA's runners.

Pull request vetters can view their responsibilities here.

Contributors can view more details about this message here.

[github-actions]

👋 Thank you for your contribution!

This pull request is from a forked repository so GitHub Actions will not be able to run CI. A maintainer will review your changes shortly and manually trigger the CI.

@maintainers Please review this PR when you have a chance and follow the instructions in the CONTRIBUTING.md file to trigger the CI.