Enabling designers, creators, and founders to plan visual projects on an AI canvas. Key features: - Ask the AI to organise your ideas, brainstorm new directions, or create a roadmap from scratch - MCP connections so you can handoff your project to an external agent to turn into something real - Your images, inspiration, and notes all on a canvas with a native agent. - Create custom widgets - ask your agent to create a habit tracker on the canvas
This looks genuinely useful — especially the ability to turn a messy canvas into something structured with AI. Curious to see how the MCP connections work in practice. Congrats on the launch!
The spatial context claim is the one I'd poke at. On our side the hard part was never generating the thing, it's that the model has no idea what the user already put next to it, so it keeps producing standalone output that doesn't belong on the board. If relative positioning actually survives the MCP handoff you've solved something most agent tools quietly skip. The custom widget bit reads like a demo feature to me, happy to be wrong.
This looks genuinely useful — especially the ability to turn a messy canvas into something structured with AI. Curious to see how the MCP connections work in practice. Congrats on the launch!
Causal
@elena_kondratenko Thank you so much. I appreciate the kind words! You can learn more about MCP set up here: https://causal.so/mcp
The spatial context claim is the one I'd poke at. On our side the hard part was never generating the thing, it's that the model has no idea what the user already put next to it, so it keeps producing standalone output that doesn't belong on the board. If relative positioning actually survives the MCP handoff you've solved something most agent tools quietly skip. The custom widget bit reads like a demo feature to me, happy to be wrong.