Enable AI Agent to push generated PNGs and Tables directly to Canvas

Customers using the AI Agent can generate insights in the form of:

• PNGs (charts, visualizations, cause maps)
• Tables (structured results)

However:

• There is no supported, documented way for the AI Agent to directly push these outputs into Canvas.
• Users must manually download images or recreate tables, then upload or rebuild them in Canvas.
• This breaks workflow continuity and reduces the usability of AI Agent outputs in collaborative analysis.

Proposed Enhancement
Introduce a native capability for AI Agents to:
1. Push generated PNGs directly into Canvas as embedded visuals.
2. Push generated tables into Canvas, ideally as:
A Canvas DataGrid (when mature), or
A structured table block that preserves rows, columns, and formatting.
This could be exposed as:

• An “Add to Canvas” / “Send to Canvas” action from the AI Agent output.
• A programmatic API available to agents and workflows.