[Code] Optimize hck_gpt/ modules

[HuckleR2003]

Milestone: v1.7.1

Target files:

• hck_gpt/chat_handler.py (~400 lines)
• hck_gpt/insights.py (~300 lines)
• hck_gpt/panel.py (~250 lines)

Changes:

• Consolidate command routing (stats, alerts, insights share similar logic)
• Remove duplicate greeting messages
• Extract hardcoded responses to config/responses.json
• Better pattern detection caching (don't recompute 7-day patterns every 30s)
• Type hints + docstrings

Performance:

• Current: insights ticker runs every 30s
• Optimize: Only compute if data actually changed
• Cache pattern results for 5min

Success:

• 950 -> 700 lines total
• Insights lag < 100ms (currently ~200ms)
• Easier to add new commands
• Polish language responses stay working