[KVR Vishnu (VISHNU0906)]

Fixes #3853

UnicodeSubstitutionConverter adds start_value to each code point and passes the sum straight to chr(). With some accepted start values (or with the default and a high input character, like U+30000) the sum goes past U+10FFFF or lands in the surrogate range, so you either get a bare chr() arg not in range error or output that blows up later on UTF-8 encode.

convert_async now checks each shifted value and raises a ValueError that names the start value, the character and the result. The constructor check stays the same, since whether a start value works depends on the prompt.

Tests cover the two cases from the issue, the surrogate start, the default start with a high input character, and the valid edges right next to the surrogate range and at U+10FFFF.

Local checks: pre-commit run --all-files, ty check pyrit tests/unit, and the full unit suite (34263 passed).

cc Roman Lutz (@romanlutz)

[Copilot]

🟡 Changes recommended

Both new parametrized tests omit required parameter and return annotations and use positional parameters, contrary to the repository style guide.

1 open issue

• Add annotations and keyword-only parameters to parametrized tests · New

What changed in this PR

This PR addresses #3853 by validating Unicode scalar values in UnicodeSubstitutionConverter, preventing out-of-range errors and surrogate output that cannot be UTF-8 encoded.

Changes:

• Rejects invalid shifted values with a descriptive ValueError, while preserving constructor validation.
• Adds regression tests for invalid combinations and valid Unicode boundaries.

File	Description
tests/​unit/​converter/​test_unicode_sub_converter.py	Tests invalid scalar rejection and valid boundary output.
pyrit/​converter/​unicode_sub_converter.py	Validates shifted code points and documents the new error behavior.

_{Review effort: Balanced}