5Claude Code Release Notes·
Claude Code 2.1.186 adds MCP CLI auth, agent-team improvements, and numerous fixes
Claude Code version 2.1.186 ships new CLI commands for authenticating MCP servers without the interactive menu (`claude mcp login/logout`), including SSH-compatible stdin redirect support. Agent-team behavior is improved: background subagents now surface permission prompts in the main session rather than auto-denying, teammate spawns inherit the leader's effort level, and subagents looping on schema validation failures now abort after 5 attempts. The release also fixes over 20 bugs spanning streaming reliability, subagent transcript handling, Chrome tab-group isolation, and session cost display for Enterprise/Team subscribers.
Related guides (3)
Related events (8)
Claude Code 2.1.203 release: major background agent reliability fixes and MCP roots integration
Anthropic released Claude Code version 2.1.203, a substantial patch addressing over 25 bugs primarily affecting background agent sessions, subagent orchestration, and worktree isolation. Key fixes include automatic recovery when daemon session tokens go stale, correct subagent state carry-over when returning to sessions, and PATH/environment variable inheritance issues on Windows. The release also adds MCP roots/list integration for working directories and improves streaming responsiveness.
Claude Code 2.1.196 release: org-default models, MCP security fix, background agent reliability improvements
Anthropic released Claude Code version 2.1.196 with a notable security fix preventing untrusted workspace .mcp.json servers from auto-spawning, and added support for organization-level default model configuration via the admin console. The release includes significant background agent reliability improvements — long-running workflows now survive process restarts on all platforms including Windows — plus a streaming idle watchdog enabled by default that aborts and retries stalled response streams after 5 minutes. Numerous bug fixes address agent session state management, MCP OAuth scope handling, PowerShell compatibility, and a /code-review token usage reduction of ~25%.
Claude Code 2.1.187 release: sandbox credentials, org model restrictions, MCP timeout fixes
Anthropic released Claude Code version 2.1.187 with a mix of security, reliability, and UX improvements. Notable additions include a sandbox.credentials setting to block credential file access in sandboxed commands, org-level model restriction enforcement in the model picker, and mouse click support in fullscreen menus. Key bug fixes address remote MCP tool calls hanging indefinitely (now abort after 5 minutes), subagent depth tracking on resume, leaked agent worktree registrations, and CJK text mojibake in certain terminals.
Claude Code 2.1.191 release: /rewind support, MCP reliability improvements, 37% CPU reduction
Anthropic shipped Claude Code version 2.1.191 with a notable set of bug fixes and performance improvements. Key additions include /rewind support for resuming conversations before a /clear, roughly 37% CPU reduction during streaming via coalesced text updates, and improved MCP server reliability with retry logic for transient network errors. Several agent-related bugs were fixed, including background agents resurrecting after being stopped and slash commands being incorrectly forwarded to background sessions.
Claude Code 2.1.199 release: multi-agent reliability and background agent fixes
Anthropic released Claude Code version 2.1.199, a patch-heavy update addressing roughly 25 bugs across multi-agent orchestration, background agent lifecycle management, SSL/TLS error handling, and streaming response reliability. Key fixes include subagents silently swallowing errors or partial work on rate-limit/server failures, a Linux daemon self-termination bug every ~50 seconds after unclean shutdown, and a macOS SSH cold-start regression from 2.1.196. The release also improves retry behavior for transient 429 errors and raises default retry counts via CLAUDE_CODE_RETRY_WATCHDOG.
Claude Code 2.1.178: Simplified agent teams, parameter-based permission rules, and subagent safety improvements
Claude Code version 2.1.178 overhauls the experimental agent teams feature by removing explicit TeamCreate/TeamDelete tools in favor of an implicit single-team model where subagents are spawned directly via the Agent tool's name parameter. A new Tool(param:value) permission syntax allows matching tool invocations by input parameter values (e.g., blocking Opus subagents specifically). A notable safety fix closes a gap where subagents could request blocked actions without classifier review before launch. The release also includes numerous bug fixes for subagent transcript visibility, auth token handling, compaction fallback, and nested .claude/ directory resolution.
Claude Code 2.1.193 adds shell classifier routing, OTel response logging, and background agent fixes
Claude Code version 2.1.193 ships several notable additions: a new autoMode.classifyAllShell setting to route all shell commands through the auto-mode classifier, an OpenTelemetry log event for assistant responses (redacted by default, with opt-in via environment variable), and live file path autocomplete in bash mode. The release also fixes multiple background agent bugs including phantom subagent spawning, spurious cancellation on backgrounding, and agent panel display issues, while improving MCP authentication reconnection on 401/403 errors.
Claude Code 2.1.198 release: GA Chrome integration, background agent PRs, AWS gateway, and extensive fixes
Anthropic shipped Claude Code version 2.1.198 with several notable capability additions: Claude in Chrome is now generally available, background agents now automatically commit and open draft PRs on task completion, and AWS (anthropicAws) is added as a gateway upstream provider. The release also includes improved subagent orchestration (extended thinking inheritance, focus mode summaries, failure reporting), a new /dataviz skill, and a large batch of bug fixes covering network resilience, macOS entitlements, session state, and UI rendering.