[Release] anthropics/claude-code v2.1.76: v2.1.76

Claude Code v2.1.76 introduces MCP elicitation support for interactive structured input, new hooks for intercepting responses, and the `/effort` slash command for model effort control. The release includes significant bug fixes for deferred tools losing schemas after compaction, plan mode re-approval issues, voice mode input handling, and Remote Control session stability. Performance improvements include optimized worktree startup, better background agent handling, and improved model fallback notifications.

Key Points

• •MCP servers can now request structured input mid-task via interactive dialogs with form fields or browser URLs using new Elicitation and ElicitationResult hooks
• •Fixed critical bug where deferred tools loaded via ToolSearch lost input schemas after conversation compaction, causing type errors on array and number parameters
• •Added `worktree.sparsePaths` setting to optimize large monorepo performance by checking out only needed directories via git sparse-checkout
• •Introduced `/effort` slash command to set model effort level and `PostCompact` hook for post-compaction actions
• •Fixed plan mode asking for re-approval after plan was already accepted, improving user experience in planning workflows
• •Resolved voice mode issues including keypresses being swallowed by permission dialogs and Windows npm installation failures
• •Improved Remote Control stability with fixes for silent session deaths, message batching, and WebSocket disconnect recovery
• •Added `-n/--name` CLI flag for setting display names at startup and session quality survey with configurable `feedbackSurveyRate`
• •Enhanced worktree startup performance by reading git refs directly and skipping redundant fetches when remote branch is available locally
• •Improved background agent behavior to preserve partial results when killed, and fixed LSP plugin server registration timing issues

Workflow Diagram

Concepts