Title: fix(base): strip markdown asterisks in pairs instead of removing all

What does this PR do?

prepare_tts_text() in the base gateway adapter used a character-class regex [*_#[]()]that stripped **every** asterisk from text, breaking math expressions like5 * 3 = 15(rendered as5 3 = 15`), pointer notation, and any other literal asterisk usage.

Fix: apply markdown-aware paired-asterisk regexes (***, **, *) before the character-class strip for remaining markdown characters, so that only formatting asterisks are removed while standalone ones survive.

Related Issue

Fixes #28010

Type of Change

• 🐛 Bug fix (non-breaking change that fixes an issue)

Changes Made

• gateway/platforms/base.py: Replaced re.sub(r'[*_#[]()]', '', text)with three paired-asterisk regexes followed by a reduced character class[_#\[\]()] that no longer includes *

How to Test

1. Set up a gateway platform with TTS enabled (e.g. Telegram voice replies)
2. Send a message containing literal asterisks like 5 * 3 = 15 or foo *bar* baz
3. Before fix: TTS reads "5 3 = 15" (asterisk stripped) and "foo bar baz" (italic stripped)
4. After fix: TTS reads "5 * 3 = 15" (standalone * preserved) while *italic* still correctly strips markers

Checklist

Code

• I've read the Contributing Guide
• My commit messages follow Conventional Commits
• I searched for existing PRs to make sure this isn't a duplicate
• My PR contains only changes related to this fix/feature
• I've run pytest tests/ -q and all tests pass
• I've added tests for my changes
• I've tested on my platform: Windows 11

Documentation & Housekeeping

• N/A — no config, docs, or schema changes