hermes - 💡(How to fix) Fix feishu: optional profile name + actual model prefix for multi-bot identification

Description

When running multiple Hermes profiles (main/daily/knowledge/dev/user-twin) with distinct Feishu bots, it is difficult to tell which bot responded and which model was actually used (after credential pool fallback). Add an optional prefix to outgoing messages.

Use Case

• 5 Feishu bots, each on its own profile
• Models can switch via credential pool fallback mid-conversation
• Need to see [profile | model] at a glance

Proposed Design

1. Config toggle — display.prefix.show_profile: true and display.prefix.show_model: true (default false for backward compat)
2. Model source — run_agent.py already stores self.model in result dict. Should pass through send() metadata
3. All platforms — prefix mechanism in BasePlatformAdapter.send(), not just feishu
4. Profile detection — from get_hermes_home() path (/profiles/<name>/)

Related Code

• gateway/platforms/feishu.py — send() method
• gateway/run.py — _run_agent() returns "model" in result dict
• run_agent.py — "model": self.model already in result dict around L15111