Opus 5, mid-conversation tool changes, and fallback mode

Patch Changes

• b4c4426: feat (provider/anthropic): support fallbacks 'default' mode, which routes safety classifier refusals to Anthropic's recommended fallback model (adds the server-side-fallback-2026-07-01 beta automatically)
• b4c4426: feat (provider/anthropic): support mid-conversation tool changes via the toolChanges system message provider option, emitting tool_addition/tool_removal content blocks and the mid-conversation-tool-changes-2026-07-01 beta
• b4c4426: feat (provider/anthropic): add claude-opus-5 model id with frontier-tier capabilities (128k output tokens, structured output, adaptive thinking, xhigh effort, sampling parameter rejection, thinking-disabled only at effort high or below)