feat: support multimodal tool outputs (text + image)

[tpirc3]

Summary

• extend FunctionCallOutput.output to support ImageContent and list[str | ImageContent]
• add shared normalization/splitting/text-fallback helpers for tool outputs
• implement provider-specific handling for tool-result images:
  • native: OpenAI Responses, Anthropic, Google, AWS standard LLM path
  • fallback to text placeholder for unsupported paths (OpenAI chat default, Mistral, realtime variants)
• add optional OpenAI chat-completions flag supports_tool_image_output to support Qwen-compatible providers that accept tool image content
• fix telemetry/span/type assumptions by converting multimodal outputs to text where string-only sinks are required

Why

This PR implements the approach proposed in #4893 for multimodal tool outputs.

Concretely, it carries the issue proposal through the shared normalization layer and provider formatters, using native image tool-output payloads where supported and explicit text fallback where not.

Resolves #4893.

Tests

• make check
• uv run pytest tests/test_tool_output_multimodal.py tests/test_chat_ctx.py tests/test_tools.py -q

[CLAassistant]

All committers have signed the CLA.