Claude Code 2.1.238 adds Concise output, idle notices, and safer session messaging
Claude Code’s August 20, 2026 release adds a built-in Concise output style, while the surrounding 2.1.238 updates improve cross-session notifications, model defaults, Remote Control, and self-hosted runners.
Editorial Team
The Authority AI Tools editorial team maintains this directory using vendor documentation, dated source checks, product changelogs, and clearly identified hands-on observations where available.
Anthropic’s latest Claude Code release adds a built-in Concise output style and continues a week of improvements to cross-session work, Remote Control, self-hosted runners, and permission feedback. The changelog labels the August 20 update 2.1.238.
Anthropic's terminal-based AI coding agent with Claude Opus 4.7, /ultrareview, Routines, /ultraplan, and 80.9% SWE-bench
Concise output style
Claude Code now includes a Concise output style. Anthropic describes it as leading with the result, skipping preamble and narration, and still doing the work thoroughly. Users can select it under Output style in /config.
The option is useful for terminal workflows where the developer wants the same underlying task execution with less explanatory framing in the visible response. It is a presentation setting, not a separate model or a promise that tool work will be reduced.
One-shot idle notices for cross-session work
The August 19 changelog entry adds notify_when_idle to cross-session SendMessage. A Claude Code session can ask another session on the same machine to send one notice the next time it becomes idle. Anthropic describes the behavior as opt-in, one-shot, and not based on polling; the entry lists macOS and Linux support.
The same week’s fixes improve feedback when cross-session delivery is refused or an inbox drops a message because of a rate limit or a full queue. Instead of reporting silent success, the sending session can now learn that the message was refused or dropped.
Model defaults and Remote Control reliability
Claude Code adds ANTHROPIC_DEFAULT_MODEL for choosing the model new sessions start with. A /model selection can still override it and persist across restarts, unlike ANTHROPIC_MODEL.
The release also improves Remote Control behavior: server-mode sessions apply the same enterprise-gateway availability check as interactive startup, and brief HTTP 403 responses from a network edge, VPN, or proxy can be tolerated for up to three minutes before a persistent block is reported. Remote sessions can also be reused after a crashed process instead of requiring a full claude remote-control restart.
Self-hosted runner and MCP changes
The August 20 entry adds a self-hosted-runner option to defer shutdown for a configured number of minutes, allowing attached sessions to keep serving before the runner parks remaining work and exits. It also adds proxy-authorization command and file options for egress proxies that need a freshly issued header on each connection.
MCP marketplace and helper changes add clearer approval boundaries. Header helpers run only in the documented install or update flow after the command is shown, and project-scoped helpers require the folder’s trust dialog to be accepted. The changelog also notes fixes for long-session memory growth and output styles drifting back to the default voice.
Sources
- Claude Code Docs — official changelog, August 20, 2026 / version 2.1.238: https://code.claude.com/docs/en/changelog
- Claude Code Docs — cross-session messaging: https://code.claude.com/docs/en/cross-session-messaging
- ClaudeDevs on X — official Concise output announcement: https://x.com/ClaudeDevs/status/2090245922685063634
Tools Mentioned in This Article
Free Resource
2026 AI Coding Tools Comparison Chart
Side-by-side comparison of features, pricing, and capabilities for every major AI coding tool.
No spam, unsubscribe anytime.
Workflow Resources
Cookbook
AI-Powered Code Review & Quality
Automate code review and enforce quality standards using AI-powered tools and agentic workflows.
Cookbook
Building AI-Powered Applications
Build applications powered by LLMs, RAG, and AI agents using Claude Code, Cursor, and modern AI frameworks.
Cookbook
Building APIs & Backends with AI Agents
Design and build robust APIs and backend services with AI coding agents, from REST to GraphQL.
Cookbook
Debugging with AI Agents
Systematically debug complex issues using AI coding agents with structured workflows and MCP integrations.
MCP Server
AWS MCP Server
Interact with AWS services including S3, Lambda, CloudWatch, and ECS from your AI coding assistant.
MCP Server
Context7 MCP Server
Fetch up-to-date library documentation and code examples directly into your AI coding assistant.
MCP Server
Docker MCP Server
Manage Docker containers, images, and builds directly from your AI coding assistant.
MCP Server
Figma MCP Server
Access Figma designs, extract design tokens, and generate code from your design files.
Frequently Asked Questions
What is Claude Code 2.1.238 adds Concise output, idle notices, and safer session messaging?
Related Articles
Google Antigravity adds IDE extensions, Gemini Enterprise access, and 2.9.1 polish
Google Antigravity’s August 20, 2026 updates add extensions for VS Code, Visual Studio, JetBrains, and Zed, bring Antigravity to eligible Gemini Enterprise plans, and improve project switching in version 2.9.1.
Read more →Product UpdatesWarp Factories brings closed-beta software factories to the SDLC
Warp’s August 18, 2026 launch introduces Warp Factories, an infrastructure layer for version-controlled cloud software factories that triage, specify, implement, review, and verify work with agents.
Read more →Product UpdatesOpenAI’s August Codex update adds Remote controls, MCP forms, and sharing
OpenAI’s August 18, 2026 Codex release improves the ChatGPT iOS experience with Codex Remote launch controls, standard MCP forms, editable approvals, and more reliable task handling.
Read more →