fix: decode help-banner test subprocess output as UTF-8

[mashraf-222]

Problem

tests/test_help_banner.py has been failing on unit-tests (windows-latest, 3.13) since PR #2014 introduced it. CI on main is red — see https://github.com/codeflash-ai/codeflash/actions/runs/25056045918 and earlier. The failure is TypeError: argument of type 'NoneType' is not iterable at assert "codeflash.ai" in result.stdout.

Root Cause

print_codeflash_banner() renders a Rich Panel that contains box-drawing characters (╭, ╮, │, ╰, etc.). The subprocess call uses subprocess.run(..., text=True) without an explicit encoding=. On Windows, text=True defaults to cp1252, which cannot decode those code points. Python's subprocess catches the UnicodeDecodeError internally and sets result.stdout = None, which is what surfaces as the confusing NoneType error at the assertion.

Fix

Pass encoding="utf-8" explicitly on both subprocess.run calls — matches the repo's existing code-style.md rule that requires encoding="utf-8" on open/read_text/write_text.

Validation

$ uv run pytest tests/test_help_banner.py -v
tests/test_help_banner.py::test_help_displays_logo PASSED         [ 50%]
tests/test_help_banner.py::test_help_short_flag_displays_logo PASSED [100%]

Tests were already passing on Linux (ubuntu-latest with default UTF-8 locale); the change does not regress them. This unblocks the required checks passed gate on PRs #1947, #1950, #1951, #1953, and every future PR touching anything in the unit-tests matrix.