osc.py: Preserve SystemExit context when re-raising ToolError in run() (B904)

[coderabbitai]

Summary

In src/rhos_ls_mcps/osc.py, the run() method of MyOpenStackShell catches SystemExit as e and raises a ToolError without chaining the original exception, which drops the SystemExit as the cause:

except SystemExit as e:
    raise ToolError(
        f"OpenStack failed {e.code}: {self.stdout.getvalue() or self.stderr.getvalue()}"
    )

This should be changed to:

except SystemExit as e:
    raise ToolError(
        f"OpenStack failed {e.code}: {self.stdout.getvalue() or self.stderr.getvalue()}"
    ) from e

Why

Without from e, the original SystemExit is silently discarded as the exception cause. Using raise ... from e preserves the full exception chain, which aids debugging. This is also flagged by Ruff rule B904 (Within an except clause, raise exceptions with raise ... from err or raise ... from None to distinguish them from errors in exception handling).

Context

Identified during review of PR #11 (Enable ruff and pre-commit). Deferred to a follow-up PR to keep #11 focused on tooling/formatting changes.

Related comment: #11 (comment)

Requested by @lpiwowar.