[Feature Request] Running-agents dropdown missing model and effort level display per subagent

Bug Description
Type: missing capability · Area: subagents / task panel

Title: Running-agents dropdown shows no model or effort level per subagent

What happened: The running-agents dropdown renders each background subagent as ◯ general-purpose <status> 1m 32s · ↓ 90.7k tokens. It shows no model and no reasoning effort. When
several subagents run concurrently at different tiers (some Sonnet, some Opus, set via the Agent tool's model parameter), there is no way to tell them apart from the panel. /agents,
/tasks, and agent view do not show it either; the only way to confirm a running agent's model is to open its transcript.

What the user said: "Is it possible to depict the model and the effort level? Next to the general purpose? So it would look something like "general-purpurse Sonnet:high""

Repro: Spawn two background subagents via the Agent tool with subagent_type: "general-purpose" and different model values (e.g. sonnet and opus). Open the running-agents dropdown.
Both rows read general-purpose with no distinguishing model or effort text.

Evidence: Claude Code 2.1.246. Statusline exposes model.display_name and effort.level, but those describe the main session only, with no per-subagent equivalent. The panel row label
is derived from the agent definition's name field, so the only workaround is defining a custom agent type whose name embeds the tier (e.g. researcher-sonnet-high). That is cosmetic,
is not validated against the frontmatter model, and does not cover call-time model overrides on general-purpose.

Cause: Not a defect. No per-subagent model or effort field is plumbed to the panel row in this version.

Environment Info

• Platform: darwin
• Terminal: Orca
• Version: 2.1.241
• Feedback ID: 667ba419-2eb2-460c-9e92-33e62929420b

Errors

[]