llm-anthropic 0.26: Claude 5 models and server-side tools

The llm-anthropic plugin has been updated to version 0.26, introducing several new features enabled by LLM 0.32. New models include claude-fable-5, claude-sonnet-5, and claude-opus-5. Server-side tools for WebSearch, WebFetch, CodeExecution, and AnthropicMCP have been added, accessible via the LLM -T interface or Python tools=. The previous -o web_search options were removed, replaced by -T WebSearch. The plugin now requires llm>=0.32, and reasoning, tool calls, tool results, and server-side tool results stream as typed events. Reasoning for llm CLI prompts displays to standard error unless --hide-reasoning/-R is passed. Extended thinking was simplified to 'thinking' and 'thinking_effort' (low, medium, high, xhigh, or max). Claude 5 models think by default; -o thinking 0 disables thinking for Sonnet 5 and Opus 5, while Fable 5 always thinks. The thinking_budget, thinking_display, and thinking_adaptive options were removed.