Expand Railway remote-ops tests beyond the current happy path

[VaibhavUpreti]

Why This Matters

tests/remote/test_ops.py covers provider resolution and one scoped status path, but not logs, fetch-logs, restart, or JSON parse failures.

Read First

• app/remote/ops.py
• tests/remote/test_ops.py

Scope

1. Add direct tests for logs(), fetch_logs(), and restart().
2. Add one test for invalid JSON output in _read_json().
3. Add one test for stderr-only log output handling in fetch_logs().

Tests / Coverage

• extend tests/remote/test_ops.py
• cover RailwayRemoteOpsProvider.logs()
• cover RailwayRemoteOpsProvider.fetch_logs()
• cover RailwayRemoteOpsProvider.restart()
• cover JSON decode failure paths

Acceptance Criteria

• remote-ops tests cover more than the status path
• parsing and stderr behavior are directly tested
• no real Railway CLI is required

[Dipxssi]

Hi @VaibhavUpreti , I would like to work on this
I plan to extend the testing for RailwayRemoteOpsProvider in tests/remote/test_ops.py by
Extending tests/remote/test_ops.py with direct tests for logs(), fetch_logs(), and restart().

Test Log Handling:
Verify that fetch_logs() correctly captures stdout.
Add a test for stderr-only output handling (e.g., when the CLI reports an error via stderr but returns 0).
Verify the formatting of combined stdout/stderr logs.

Test Parsing Robustness:
Add a test case for invalid JSON output in _read_json() to ensure it raises a descriptive RemoteOpsError.
Mocking: Ensure all tests use unittest.mock.patch for subprocess.run and shutil.which so they remain independent of the actual Railway CLI installation.

Please assign this to me , Thanks!