Fetching from the wire…
Public story · 2026-09-12 · high
It targets the seconds lost switching between sessions to see what each one is asking.
Why now: It's documented as teams increasingly run several Claude Code sessions in parallel and need a faster way to unstick them.
Claude Code's agent view lets you resolve a waiting session without leaving the list, per the agent view docs. Arrow to a row and press Space to open that session's latest output or its pending question. Type a reply and hit Enter to move to the next one.
Running many agents in parallel was never the slow part. It's the six questions each one stops to ask, four seconds to answer once you've found the session asking it.
The agent view already lists every dispatched session on one screen and flags which ones are waiting on input. The peek panel adds the missing step, answering from inside that same list rather than opening a separate window to see what a session asked.
The docs don't say how many sessions the panel handles before the list gets hard to scan. They also don't say whether it ranks waiting sessions by anything other than order.
This is a UI change, not a model change. Nothing about how Claude reasons is different, only how fast you get it unstuck. Watch whether a later version ranks waiting questions by importance instead of listing them in the order they arrived.
Each link below shares sources, entities, or timing with this story.
xAI launched Grok Build on May 14. With that, every major AI lab now ships a coding agent that lives in your terminal. The competition isn't "can we build one" anymore. That question is settled. The lineup: Anthropic has Claude Code. OpenAI has Codex CLI. Google has Gemini CLI...
It now dedupes local CLAUDE.md files against checked-in ones, proposes trimming content it can derive from the codebase, ranks unused skills, MCP servers, and plugins by context cost, and flags slow hooks. Background subagents run by default and, when launched from claude agen...
2.1.224 shipped SendMessage and ListAgents as first-class tools so separate sessions on macOS/Linux address each other by name, governed by crossSessionInbound and dialogExpiry. Auto mode routes inbound message content through the permission classifier before dispatch, treatin...
The TypeScript project compiles a repository into an explorable, searchable graph you can ask questions against, with the explicit framing "graphs that teach beat graphs that impress." It targets Claude Code, Codex, Cursor, and Copilot. Directly relevant if your agent currentl...
Three separate Anthropic changes over about two weeks point the same direction, and none of them announced themselves as a strategy. Claude Code 2.1.238 added claude self-hosted-runner --defer-shutdown-max-min, which keeps serving attached sessions on SIGTERM, parks whatever's...
Claude Code 2.1.232 shipped August 13 with a default flipped underneath everyone. Subagent forking is now on: a spawned subagent inherits the full parent conversation and shares its prompt-cache prefix instead of starting cold. Nested spawn depth defaults to 3. Practitioner wr...
MindPattern daily
One email a day at 7 AM. Sources and a take on every story. Unsubscribe anytime.