Describe seconv MCP container track defaults accurately

[Blackspirits]

Summary

Upstream PR SubtitleEdit#14129 describes convert_subtitle.trackNumbers as:

Default: every text track.

That is narrower than the actual shared converter behavior. With no track filter, the container loader also processes supported image subtitle tracks:

• Matroska PGS;
• Matroska/MP4 VobSub;
• other image-based paths handled by the shared loader.

Those tracks can trigger OCR unless timeCodesOnly is enabled. An MCP client relying on the current description can therefore incorrectly assume that an unfiltered container conversion is text-only and cheap.

This change updates only the tool parameter description to state that the default is every supported subtitle track, including supported image tracks that may require OCR.

Validation

• based exactly on upstream PR seconv: add seconv mcp, a Model Context Protocol server over stdio SubtitleEdit/subtitleedit#14129 HEAD ad51c0ae9f5e8942f281e4acd82090974f9b8847
• documentation/schema metadata only; no conversion behavior changes
• full solution CI pending

AI assistance: ChatGPT was used to compare MCP tool descriptions against the shared container loader.