ops-sync: Add SSH timeout, improve error detection and reporting

[khamer]

Summary

• Add OPS_SSH_CONNECT_TIMEOUT config (default 10s) to detect hung SSH connections during ops sync
• Capture SSH stderr to temp log and report connection failures with helpful context (firewall, VPN, bastion host suggestions)
• Check for empty database dumps before dropping database to avoid data loss on failed syncs
• Fix shell portability: replace seq with printf for padding, use arithmetic operators for comparisons
• Remove OPS_TEST_MODE (undocumented, no longer used)
• Enforce proper exit codes: unknown commands return 1, missing remote host exits early, ops sync outside project exits 1
• Redirect error messages to stderr throughout

What changed

ops-sync reliability:

• SSH connection timeout now detected and reported clearly
• Database dump failures now surface SSH stderr instead of silently failing
• Empty dumps are rejected before destructive DROP DATABASE
• Added validation that OPS_PROJECT_REMOTE_HOST is configured

cmd-run error handling:

• Unknown commands now print error and return exit code 1
• Preserves exit code from actual command execution

Shell improvements:

• Portable padding using printf instead of macOS-incompatible seq
• Arithmetic comparisons use (( )) instead of [[ ]]

🤖 Generated with Claude Code

[khamer]

Requesting Copilot, and going to run this branch locally for a bit.

[Copilot]

Copilot encountered an error and was unable to review this pull request. You can try again by re-requesting a review.

[Copilot]

Copilot review overview

🟡 Changes recommended

Database importer failures are currently masked when SSH succeeds.

Review effort: Balanced
Findings: 1

Open (1)

• Propagate local importer failures after SSH pipeline execution · New