Problem or Motivation
There is no single command to view the current session state. Users need to run /model, /auth, and mentally track token consumption separately. Key information like cumulative token usage, active provider, default region, session ID, and working directory are scattered or unavailable.
This is especially useful for:
- Monitoring costs during long sessions
- Debugging provider/model configuration issues
- Quickly verifying which credentials and region are active before a deployment
Proposed Solution
Add a /status command that displays a concise overview of the current session:
╭─ Session Status ─────────────────────────────────────────────╮
│ Session: e91c5730 (resumed) │
│ Provider: DashScope (qwen3.7-max) │
│ Region: cn-beijing │
│ CWD: /Users/dev/infra │
│ │
│ Token Usage (this session): │
│ Input: 12,450 tokens │
│ Output: 3,280 tokens │
│ Cache: 8,200 tokens (read) │
│ Total: 15,730 tokens │
│ │
│ Turns: 7 / 100 │
│ Context: 45% used (58k / 128k) │
╰──────────────────────────────────────────────────────────────╯
Key fields:
- Session ID and whether it was resumed
- Provider + model currently active
- Default region for cloud operations
- Cumulative token usage (input, output, cache read/create) for the session
- Turn count vs max turns
- Context window utilization (current tokens / window size)
Related Area
CLI / Interactive Mode
Additional Context
The renderer already tracks per-message token usage (_format_token_count in renderer.py), and ContextManager.get_usage() provides context utilization data. The main work is accumulating session-level totals and wiring them into a new slash command.
Problem or Motivation
There is no single command to view the current session state. Users need to run
/model,/auth, and mentally track token consumption separately. Key information like cumulative token usage, active provider, default region, session ID, and working directory are scattered or unavailable.This is especially useful for:
Proposed Solution
Add a
/statuscommand that displays a concise overview of the current session:Key fields:
Related Area
CLI / Interactive Mode
Additional Context
The renderer already tracks per-message token usage (
_format_token_countinrenderer.py), andContextManager.get_usage()provides context utilization data. The main work is accumulating session-level totals and wiring them into a new slash command.