fix: display active model in /status command

[GautamKumarOffical]

Summary

The /status command was reading the model from the config file, but when users switch models using /model without --global, the model wasn't being persisted to the config. This caused /status to show the old model instead of the active one.

Now /status checks _session_model_overrides first (where /model stores session-specific switches), then falls back to the session database, and finally to the config file.

Changes

• Added check for _session_model_overrides in _handle_status_command
• Session-specific model overrides now take precedence over config defaults
• No breaking changes - fully backward compatible

Test plan

• Start hermes gateway
• Use /model deepseek/deepseek-v4-pro to switch model
• Use /status to verify the new model is displayed
• Use /model --global openai/gpt-4o to switch globally
• Use /status to verify the global model is displayed

Closes #48715

[alt-glitch]

Duplicate of #11079 — both fix #48715 by making /status read _session_model_overrides (the session-specific model switch) before falling back to config, so the displayed model stays accurate after a /model switch. #11079 (open, earlier) applies the same fix in gateway/run.py::_handle_status_command; this PR applies it in gateway/slash_commands.py::_handle_status_command (the handler now lives in both files on main). Same mechanism, same issue — consolidating on the earlier open PR.

[tonydwb]

Code Review Summary

Verdict: Approved

• 1 file changed, +7/-0 lines — displays active model in status command output
• Small, targeted fix; no security or test concerns

---

Reviewed by Hermes Agent

[GautamKumarOffical]

Closing — marked duplicate of #11079.