Claude-Code v2.1.80 adds rate limit monitoring, plugin improvements, and memory optimizations

What's new in Claude-Code v2.1.80
Claude-Code v2.1.80 includes several practical improvements for developers using AI coding agents, focusing on monitoring, plugin management, and performance optimizations.
Key features and improvements
- Rate limit monitoring: Added
rate_limitsfield to statusline scripts for displaying Claude.ai rate limit usage with 5-hour and 7-day windows showingused_percentageandresets_at - Plugin marketplace source: Added
source: 'settings'option to declare plugin entries inline insettings.json - CLI tool detection: Added CLI tool usage detection to plugin tips, complementing existing file pattern matching
- Effort level control: Added
effortfrontmatter support for skills and slash commands to override the model effort level when invoked - MCP server channels: Added
--channelsflag (research preview) allowing MCP servers to push messages into your session - Memory optimization: Reduced memory usage on startup in large repositories, saving approximately 80 MB on repositories with 250,000 files
Bug fixes
- Fixed
--resumedropping parallel tool results - sessions with parallel tool calls now restore alltool_use/tool_resultpairs instead of showing[Tool result missing]placeholders - Fixed voice mode WebSocket failures caused by Cloudflare bot detection on non-browser TLS fingerprints
- Fixed 400 errors when using fine-grained tool streaming through API proxies, Bedrock, or Vertex
- Fixed
/remote-controlappearing for gateway and third-party provider deployments where it cannot function - Fixed
/sandboxtab switching not responding to Tab or arrow keys - Fixed managed settings (
enabledPlugins,permissions.defaultMode, policy-set env vars) not being applied at startup whenremote-settings.jsonwas cached from a prior session
UI improvements
- Improved responsiveness of
@file autocomplete in large git repositories - Improved
/effortto show what auto currently resolves to, matching the status bar indicator - Improved
/permissions- Tab and arrow keys now switch tabs from within a list - Improved background tasks panel - left arrow now closes from the list view
- Simplified plugin install tips to use a single
/plugin installcommand instead of a two-step flow
📖 Read the full source: GitHub Claude-Code
👀 See Also

Claude Code developer acknowledges adaptive thinking flaw, provides workaround
Boris Charny, creator of Claude Code, confirmed a flaw in the adaptive thinking feature that causes performance degradation. Users experiencing issues even with effort=high settings can use CLAUDE_CODE_DISABLE_ADAPTIVE_THINKING=1 as an interim workaround.

Anthropic releases free educational curriculum including Claude Code and MCP Mastery courses
Anthropic has made its entire educational curriculum available for free, including courses on Claude Code, MCP Mastery, API usage, and AI Fluency. The curriculum is described as university-level and provides structured learning compared to random tutorials.

Hy3 LLM Tops OpenRouter Rankings: Cheapest Model or Something Else?
Hy3 preview, a Tencent open-source LLM, surged to the top of OpenRouter's model rankings by token usage, surpassing Claude and DeepSeek V4 Flash. Priced at $0.066/1M input tokens, it's the cheapest major model, but benchmarks show quality far below leaders.

DeepSeek Paid API Uses Prompts for Training — What OpenClaw Users Need to Know
DeepSeek's official API logs prompts for training, even on paid tiers. Gemini only logs on free AI Studio. OpenClaw now defaults to DeepSeek V4 Flash — beware when processing personal data.