fix(gateway): read /status token usage from SQLite

[wgd753]

What does this PR do?

This fixes a gateway /status regression where the command can show Tokens: 0 even when the same session already has substantial token usage persisted in SQLite state.db.

Root cause:

• gateway token accounting is now written to SQLite / SessionDB
• /status still renders session_entry.total_tokens from SessionStore / sessions.json
• that mirror is no longer kept in sync, so it can drift to 0

This PR fixes the bug by making /status prefer the authoritative SQLite session stats first, while keeping SessionEntry.total_tokens as a fallback when no DB row exists. That matches the current source of truth and avoids reintroducing another mirrored accounting path that can drift again.

Related Issue

Fixes #5960
Related: #11087

Type of Change

• 🐛 Bug fix (non-breaking change that fixes an issue)
• ✨ New feature (non-breaking change that adds functionality)
• 🔒 Security fix
• 📝 Documentation update
• ✅ Tests (adding or improving test coverage)
• ♻️ Refactor (no behavior change)
• 🎯 New skill (bundled or hub)

Changes Made

• gateway/run.py
  • add _status_total_tokens() helper
  • read input_tokens, output_tokens, cache_read_tokens, and cache_write_tokens from self._session_db.get_session(session_id)
  • make /status display the SQLite-derived total instead of session_entry.total_tokens when DB stats are available
• tests/gateway/test_status_command.py
  • seed the mocked session DB with get_session.return_value = None by default
  • add a regression test proving /status prefers SQLite-backed token counts over a stale SessionStore total

How to Test

1. Reproduce on current main
   • run Hermes through the gateway
   • have a longer conversation so the session accumulates token usage
   • run /status
   • observe that it can report Tokens: 0 while the same session has non-zero usage in state.db
2. Apply this branch and run:
   • pytest tests/gateway/test_status_command.py -q
3. Verify the focused regression path:
   • python -m py_compile gateway/run.py tests/gateway/test_status_command.py
   • git diff --check
   • confirm /status now reports the SQLite-backed total instead of 0

Checklist

Code

• I've read the Contributing Guide
• My commit messages follow Conventional Commits (fix(scope):, feat(scope):, etc.)
• I searched for existing PRs to make sure this isn't a duplicate
• My PR contains only changes related to this fix/feature (no unrelated commits)
• I've run pytest tests/ -q and all tests pass
• I've added tests for my changes (required for bug fixes, strongly encouraged for features)
• I've tested on my platform: Ubuntu 24.04 / Linux

Documentation & Housekeeping

• I've updated relevant documentation (README, docs/, docstrings) — or N/A
• I've updated cli-config.yaml.example if I added/changed config keys — or N/A
• I've updated CONTRIBUTING.md or AGENTS.md if I changed architecture or workflows — or N/A
• I've considered cross-platform impact (Windows, macOS) per the compatibility guide — or N/A
• I've updated tool descriptions/schemas if I changed tool behavior — or N/A

Screenshots / Logs

Focused validation run:

$ pytest tests/gateway/test_status_[command.py](https://command.py/) -q
10 passed

Additional checks:

$ python -m py_compile gateway/[run.py](https://run.py/) tests/gateway/test_status_[command.py](https://command.py/)
# passed

$ git diff --check
# passed
```# Screenshots / Logs

Focused validation run:
 `$ pytest tests/gateway/test_status_[command.py](https://command.py/) -q`
10 passed

[alt-glitch]

Related: PRs #13404 and #5989 also fix this same /status token display regression (#5960). One of these three should be chosen.

[teknium1]

Fixed on main in 7abc9ce via #17158 — essentially your approach (inline SessionDB.get_session() + sum columns in _handle_status_command). Closing as superseded. Thanks for the fix.