agentleFS
vchelaruvchelaru/Gum/.claude/skills/gum-tool-output/SKILL.md

gum-tool-output

Skill659 starsChanged 2 months agoMITView on GitHub

Gum Output tab. Triggers: IOutputManager, MainOutputViewModel, GuiCommands.PrintOutput, adding output/error messages.

The file

Files
SKILL.md

Gum Tool Output System Reference

Architecture

IOutputManager (Tools/Gum.Presentation/Managers/MainOutputViewModel.cs):

  • AddOutput(string) — appends a timestamped line
  • AddError(string) — appends a timestamped line prefixed with "ERROR: " (two spaces), then raises ErrorAdded
  • ErrorAdded — event MainOutputPlugin subscribes to, selecting the Output tab so an error is seen without a dialog

MainOutputViewModel — implements IOutputManager. Stores all output as a single OutputText string (not a list). Registered as a singleton in DI, aliased as both MainOutputViewModel and IOutputManager.

MainOutputPlugin (Tools/Gum.Presentation/Plugins/InternalPlugins/Output/) — CorePriorityPlugin that creates the Output tab at TabLocation.RightBottom.

How to Write Output

Inject IOutputManager and call AddOutput or AddError directly — this is the standard approach. Callers must already be on the UI thread.

IGuiCommands.PrintOutput is a thin wrapper around AddOutput that marshals to the UI dispatcher — use this when calling from a background thread. There is no PrintError equivalent in IGuiCommands; callers needing AddError from a background thread must dispatch manually.

Non-Obvious Behaviors

  • Buffer cap: OutputText is capped at 50,000 chars. When exceeded, it is trimmed to the last 25,000 chars — oldest output is silently discarded.
  • No dispatcher in IOutputManager: AddOutput/AddError write directly to the OutputText property with no thread marshaling. Calling from a background thread will throw.
  • Auto-scroll: The TextBox in the view uses TextBoxAutoScroll.AutoScrollToEnd="True" — new output scrolls into view automatically.
  • AddError steals the tab: it selects the Output tab, so it is the wrong call for a routine or repeated message. Use AddOutput unless the user genuinely needs to look now.
  • Errors can arrive before the tab exists: the plugin catalog reports skipped plugin assemblies before any plugin's StartUp, so MainOutputPlugin.StartUp checks MainOutputViewModel.HasErrors as well as subscribing to ErrorAdded.
  • GUM_ECHO_OUTPUT=1 mirrors every Output line to stderr, the way to read Output from an unattended --exit-after run.
  • No per-line color: severity is carried by the "ERROR: " text prefix. The view binds one TextBox to one concatenated OutputText string, so coloring a single line needs the model changed to a line collection first.

Key Files

FilePurpose
Tools/Gum.Presentation/Managers/MainOutputViewModel.csIOutputManager interface + MainOutputViewModel implementation
Gum/Commands/GuiCommands.csPrintOutput — dispatcher-safe wrapper
Tools/Gum.Presentation/Commands/IGuiCommands.csPrintOutput declaration
Tools/Gum.Presentation/Plugins/InternalPlugins/Output/MainOutputPlugin.csRegisters the Output tab
Tool/Gum.Avalonia/Panels/ToolPanelViews.cs (OutputView)Avalonia Output tab view; WPF twin is Gum/Plugins/InternalPlugins/Output/MainOutputPluginView.xaml
Tools/Gum.Presentation/Services/GumCoreServiceCollectionExtensions.csDI registration of MainOutputViewModel as IOutputManager

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.