claude-code - 💡(How to fix) Fix Naming feedback: "agent view" conflates with the existing "agent" concept

Problem

The new agent view (v2.1.139, May 11) is described in the changelog as:

"Added agent view (Research Preview): a single list of every Claude Code session — running, blocked on you, or done."

But "agent" already has a specific, distinct meaning elsewhere in Claude Code:

• The Agent tool / subagent_type parameter
• .claude/agents/*.md definitions
• The Claude Agent SDK

In those contexts, an agent is a configured role — a system prompt + tools + model + permissions. It's the thing, not the run.

What "agent view" actually shows is the list of sessions — the changelog itself calls them sessions in the same sentence ("a single list of every Claude Code session"). The in-app hint reinforces the split: "Press the left arrow from any session or run claude agents from the terminal to open agent view."

So within one feature, the same concept is called both "session" (in description) and "agent" (in the view name + CLI command). And in the wider product, "agent" already means something else entirely.

Suggested rename

Either of these would disambiguate:

• Session view / claude sessions — most precise; already how the changelog describes the contents
• Chats / claude chats — friendlier; matches how users describe the thing ("my other chat")

"Agent" should stay reserved for the configured-role concept (Agent tool, .claude/agents/, SDK).

Impact

Minor on its own, but it's the kind of naming choice that compounds — every doc page, error message, and community post that says "agent" now has to clarify which kind. Easier to fix while the feature is still in Research Preview.