[Feature]: Per-session token usage breakdown by model

### Prerequisites

- [x] I will write this issue in English (see our [Language Policy](https://github.com/code-yeongyu/oh-my-opencode/blob/dev/CONTRIBUTING.md#language-policy))
- [x] I have searched existing issues and discussions to avoid duplicates
- [x] This feature request is specific to oh-my-opencode (not OpenCode core)
- [x] I have read the [documentation](https://github.com/code-yeongyu/oh-my-opencode#readme) or asked an AI coding agent with this project's GitHub URL loaded and couldn't find the answer

### Problem Description

When using multi-model orchestration with `ultrawork` and category-based delegation, there's no way to see how many tokens each model consumed during a session. The TUI only shows total token count for the main session, without breaking down by model (e.g., DeepSeek V4 Flash vs GLM-5.1 vs Qwen 3.6 Plus).
## Use Case
As a user with a tiered model setup (cheap models for coding, expensive models for orchestration/review), I need to:
1. Understand cost distribution per session — how much did the orchestrator vs sub-agents cost
2. Verify that cheap models are handling most of the work
3. Detect if an expensive model is being over-used unexpectedly

### Proposed Solution

A session-end summary (toast or panel) showing:
Session Token Summary:
Sisyphus (DeepSeek V4 Flash) 45,230 tokens $0.02
Momus (GLM-5-Turbo) 12,100 tokens $0.05
Sisyphus-Junior (Qwen 3.6 Plus) 89,500 tokens $0.04
──────────────────────────────────────────────
Total 146,830 tokens $0.11
This could leverage the existing `session.idle` event and toast notification system.
## Alternatives Considered
- Checking provider dashboards (fragmented across 3+ platforms, no unified view)
- Parsing log files (not user-friendly)

### Alternatives Considered

_No response_

### Doctor Output (Optional)

```shell

```

### Additional Context

_No response_

### Feature Type

New Agent

### Contribution

- [ ] I'm willing to submit a PR for this feature
- [ ] I can help with testing
- [ ] I can help with documentation