# LibTmux.Mcp.PaneSnapshot

- **Module:** LibTmux.Mcp
- **Package:** LibTmux.Mcp
- **Language:** .NET
- **Kind:** struct
- **Source:** https://github.com/libtmux/libtmux-dotnet/blob/320dc64f4b8b7815842471327a5e6b84a1499bf8/src/LibTmux.Mcp/Results/Observations.cs#L17
- **Page:** https://libtmux.org/en/dotnet/latest/mcp/reference/libtmux-mcp-panesnapshot/

A pane's content and its screen state, read together.

One call rather than a capture followed by an info read: the two would be
separated by however long the model took to ask again, and a pane can move
in between, which is how a cursor ends up described against the wrong text.