[Release] anthropics/claude-code v2.1.186: v2.1.186

Claude Code v2.1.186 introduces CLI authentication for MCP servers, enhanced workflow filtering, and improved agent team management. Key fixes address streaming failures after sleep, subagent transcript issues, and permission prompt handling. The release also improves skill metadata flexibility, memory management, and background session behavior with better user control over tool permissions.

Key Points

• •Added `claude mcp login/logout <name>` commands for CLI-based MCP server authentication with SSH support via `--no-browser` flag
• •Implemented status filtering in `/workflows` agent detail view and new 'Skills' section in `/plugin` Installed tab
• •Fixed critical streaming bug causing 'Content block not found' or JSON parse errors after machine sleep
• •Changed bash commands (!) to automatically trigger Claude responses; disable with `"respondToBashCommands": false` setting
• •Improved permission prompts: background subagents now surface permission dialogs in main session instead of auto-denying, with agent identification
• •Fixed subagent transcript scroll position bleeding and background session recap duplication issues
• •Added `teammateMode: "iterm2"` setting for agent teams with auto-detection fallback warning
• •Enhanced skill metadata parsing to accept kebab-case, snake_case, and camelCase formats with graceful fallback for malformed YAML
• •Fixed Workflow `agent({schema})` subagents looping infinitely on validation failures; now aborts after 5 attempts
• •Improved `claude mcp get/remove` commands with typo suggestions and long server list truncation

Workflow Diagram

Concepts